{ "total": 839, "topics": [ { "id": 180381, "title": "Looking for a Team Megathread", "comment_count": 221, "views": 0, "votes": 48 }, { "id": 195170, "title": "New data file available: train_drug.csv", "comment_count": 57, "views": 0, "votes": 116 }, { "id": 181436, "title": "Welcome to the MOA Prediction Challenge!", "comment_count": 63, "views": 0, "votes": 71 }, { "id": 180383, "title": "New to Machine Learning or to Kaggle? Check this out.", "comment_count": 26, "views": 0, "votes": 29 }, { "id": 201510, "title": "1st Place Winning Solution - Hungry for Gold", "comment_count": 30, "views": 0, "votes": 296 }, { "id": 196913, "title": "Some tips to avoid overfitting", "comment_count": 129, "views": 0, "votes": 285 }, { "id": 184005, "title": "Competition Insights", "comment_count": 137, "views": 0, "votes": 500 }, { "id": 233838, "title": "Please share drug names also ", "comment_count": 1, "views": 0, "votes": 4 }, { "id": 198602, "title": "Probability Calibration", "comment_count": 6, "views": 0, "votes": 32 }, { "id": 202256, "title": "2nd Place Solution - with 1D-CNN (Private LB: 0.01601) ", "comment_count": 37, "views": 0, "votes": 217 }, { "id": 190467, "title": "TabNet vs NODE", "comment_count": 15, "views": 0, "votes": 49 }, { "id": 312975, "title": "PyTorch newly added built-in CrossEntropyLoss function with Label Smoothing", "comment_count": 0, "views": 0, "votes": 3 }, { "id": 195378, "title": "DeepInsight: Transforming MoA Tabular Data to Images for CNNs", "comment_count": 38, "views": 0, "votes": 135 }, { "id": 200540, "title": "[Update]Augmentation is all you need", "comment_count": 32, "views": 0, "votes": 82 }, { "id": 200525, "title": "Congratulations to the final winners!", "comment_count": 31, "views": 0, "votes": 79 }, { "id": 201437, "title": "Feature Engineering: Independent vs. Dependent FE's", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 191621, "title": "A list of tips", "comment_count": 98, "views": 0, "votes": 167 }, { "id": 192211, "title": "(L.B.- 0.01839) Single NN Model---Useful Links", "comment_count": 142, "views": 0, "votes": 222 }, { "id": 200609, "title": "Post-Processing Based on Drug Classification (15th place)", "comment_count": 11, "views": 0, "votes": 23 }, { "id": 190175, "title": "Fixing \"Submission CSV Not Found\" and similar errors", "comment_count": 10, "views": 0, "votes": 19 }, { "id": 183377, "title": "Summary of My Novel Ideas ", "comment_count": 60, "views": 0, "votes": 280 }, { "id": 235514, "title": "Tribute to tmp and its 1D-CNN", "comment_count": 0, "views": 0, "votes": 13 }, { "id": 235679, "title": "What is the best way to classify Chemical drugs (Chemical names. Generic names. Trade names). ", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 194369, "title": "What stacking methods for this competition?", "comment_count": 17, "views": 0, "votes": 35 }, { "id": 200832, "title": "CV LB Gap Explained", "comment_count": 27, "views": 0, "votes": 97 }, { "id": 190647, "title": "EDA and Feature Importance Findings", "comment_count": 13, "views": 0, "votes": 55 }, { "id": 206242, "title": "MoA: Shakeup scatterplot", "comment_count": 6, "views": 0, "votes": 13 }, { "id": 207826, "title": "Annotated dataset with gene symbols in a single compressed csv file", "comment_count": 1, "views": 0, "votes": 3 }, { "id": 200614, "title": "3rd Place Public - We Should Have Trusted CV - 118th Private", "comment_count": 55, "views": 0, "votes": 130 }, { "id": 195067, "title": "TabNet with TF 2?", "comment_count": 11, "views": 0, "votes": 5 }, { "id": 196352, "title": "Combining Dropout and BatchNormalization", "comment_count": 3, "views": 0, "votes": 16 }, { "id": 208451, "title": "MoA Competition - Extensive blog explanation", "comment_count": 2, "views": 0, "votes": 12 }, { "id": 200545, "title": "Silver medal in very first competition!!! (A noob first Kaggle Experience)", "comment_count": 26, "views": 0, "votes": 32 }, { "id": 200808, "title": "4th Place Solution", "comment_count": 15, "views": 0, "votes": 87 }, { "id": 199078, "title": "Ensembling week!!!", "comment_count": 45, "views": 0, "votes": 21 }, { "id": 207259, "title": "Link to repo with mapping file for gene ID", "comment_count": 3, "views": 0, "votes": 13 }, { "id": 195803, "title": "How to Kaggle?", "comment_count": 6, "views": 0, "votes": 13 }, { "id": 204685, "title": "[Late Sub, Private 0.01599] Averaging Various Stacking Models", "comment_count": 8, "views": 0, "votes": 35 }, { "id": 200992, "title": "8th place solution | From multilabel to multiclass", "comment_count": 28, "views": 0, "votes": 61 }, { "id": 200509, "title": "Tribute to Yirun Zhang", "comment_count": 6, "views": 0, "votes": 93 }, { "id": 200561, "title": "Thank you all! First time competition, silver medal! My blending nn+tabnet+resnet", "comment_count": 3, "views": 0, "votes": 11 }, { "id": 200985, "title": "My First Competition Medal!", "comment_count": 3, "views": 0, "votes": 18 }, { "id": 200387, "title": "Please unanonymize genes, drugs and cell lines after the deadline ", "comment_count": 3, "views": 0, "votes": 13 }, { "id": 202612, "title": "Few Beginner Friendly Tips!", "comment_count": 2, "views": 0, "votes": 8 }, { "id": 202620, "title": "A competition expert now 😄!!", "comment_count": 15, "views": 0, "votes": 21 }, { "id": 201565, "title": "26th Place Solution / Source Code", "comment_count": 11, "views": 0, "votes": 12 }, { "id": 202615, "title": "Finally I received the first silver🥈🎉🎉", "comment_count": 5, "views": 0, "votes": 14 }, { "id": 201331, "title": "When will the medals be assigned ?", "comment_count": 17, "views": 0, "votes": 31 }, { "id": 201051, "title": "Layer Normalization - 36th place", "comment_count": 10, "views": 0, "votes": 36 }, { "id": 202457, "title": "Guessing game: When will results be out? ", "comment_count": 6, "views": 0, "votes": 3 }, { "id": 202078, "title": "3rd source code(To fulfill requirement of prize)", "comment_count": 0, "views": 0, "votes": 20 }, { "id": 195195, "title": "Drug and MultiLabel Stratification Code", "comment_count": 134, "views": 0, "votes": 278 }, { "id": 202125, "title": "MoA solution - Tried to keep it simple ", "comment_count": 2, "views": 0, "votes": 9 }, { "id": 201729, "title": "To smooth or not to smooth,... Not to smooth (three models to obtain position 20th, 0.01607)", "comment_count": 9, "views": 0, "votes": 17 }, { "id": 198945, "title": "Load models / model weights in R/Keras", "comment_count": 21, "views": 0, "votes": 13 }, { "id": 201465, "title": "MoA: Approaches to top Private NBs & Discussions", "comment_count": 5, "views": 0, "votes": 12 }, { "id": 200784, "title": "7th place solution", "comment_count": 30, "views": 0, "votes": 69 }, { "id": 200736, "title": "1st place solution-Summary", "comment_count": 26, "views": 0, "votes": 90 }, { "id": 201318, "title": "When are medals reflected in the profile?", "comment_count": 6, "views": 0, "votes": 25 }, { "id": 200680, "title": "Denoising Autoencoder (DAE) - Public LB 1830", "comment_count": 11, "views": 0, "votes": 42 }, { "id": 200533, "title": "5th place solution [Updated]", "comment_count": 26, "views": 0, "votes": 102 }, { "id": 180536, "title": "List of Solution Methods for Multi-Label Classification", "comment_count": 10, "views": 0, "votes": 92 }, { "id": 200338, "title": "Public 2nd/Private 560th (Private 0.01609 Silver Solution)", "comment_count": 49, "views": 0, "votes": 168 }, { "id": 200717, "title": "Best single model - final", "comment_count": 24, "views": 0, "votes": 15 }, { "id": 200585, "title": "Private 14th / Public 24th place solution! ", "comment_count": 36, "views": 0, "votes": 75 }, { "id": 201254, "title": "Threshold Selection for Transferring Confident Samples", "comment_count": 2, "views": 0, "votes": 12 }, { "id": 202487, "title": "My private score doesn't change for my submission", "comment_count": 4, "views": 0, "votes": -2 }, { "id": 200644, "title": "Will the drugs/gene/cell lines names be released?", "comment_count": 7, "views": 0, "votes": 5 }, { "id": 200998, "title": "Single Model - Public LB 1777 - WOW!", "comment_count": 9, "views": 0, "votes": 41 }, { "id": 200616, "title": "Things I have learned", "comment_count": 6, "views": 0, "votes": 23 }, { "id": 200535, "title": "Publishing high scoring notebooks should be stopped at least one week before the deadline", "comment_count": 29, "views": 0, "votes": 27 }, { "id": 200570, "title": "Can't believe it!", "comment_count": 27, "views": 0, "votes": 86 }, { "id": 200768, "title": "Make it simpler", "comment_count": 9, "views": 0, "votes": 49 }, { "id": 200798, "title": "Public 46th / Private 34th Solution", "comment_count": 3, "views": 0, "votes": 34 }, { "id": 200530, "title": "MoA Shake-up Plots", "comment_count": 17, "views": 0, "votes": 47 }, { "id": 200733, "title": "Top 1% write up", "comment_count": 11, "views": 0, "votes": 21 }, { "id": 200492, "title": "Don't share please", "comment_count": 2, "views": 0, "votes": 25 }, { "id": 200552, "title": "40th place solution | MLPs, ResNet, TabNet & Clustering", "comment_count": 11, "views": 0, "votes": 28 }, { "id": 200536, "title": "Finally a Gold.... WOHOOOO !!", "comment_count": 18, "views": 0, "votes": 32 }, { "id": 200845, "title": "Trust Your CV - Caveats", "comment_count": 13, "views": 0, "votes": 8 }, { "id": 201089, "title": "another way to use CNN", "comment_count": 1, "views": 0, "votes": 8 }, { "id": 200881, "title": "About my feature engineering with good results", "comment_count": 5, "views": 0, "votes": 15 }, { "id": 200610, "title": "My First Medal !!!!!!!", "comment_count": 9, "views": 0, "votes": 13 }, { "id": 200679, "title": "30th place solution", "comment_count": 3, "views": 0, "votes": 17 }, { "id": 200575, "title": "Simple Preliminary 71st Place Solution Explained", "comment_count": 16, "views": 0, "votes": 16 }, { "id": 200630, "title": "27th place - Trusting CV always works.", "comment_count": 13, "views": 0, "votes": 19 }, { "id": 200249, "title": "Public LB and potential shakeup...", "comment_count": 46, "views": 0, "votes": 24 }, { "id": 200508, "title": "Lottery ? ", "comment_count": 19, "views": 0, "votes": 25 }, { "id": 200779, "title": "How much blend gave you comparing to your best model ? (And to worst)", "comment_count": 9, "views": 0, "votes": 2 }, { "id": 200521, "title": "53rd place in a nutshell", "comment_count": 18, "views": 0, "votes": 25 }, { "id": 200596, "title": "Post-Processing Success?", "comment_count": 11, "views": 0, "votes": 13 }, { "id": 200593, "title": "Offline Pseudo labelling VS online pseudo labelling", "comment_count": 17, "views": 0, "votes": 11 }, { "id": 200795, "title": "35th place solution", "comment_count": 2, "views": 0, "votes": 16 }, { "id": 200605, "title": "Why did preprocessing of train+test work well? Wasn't it Leakage?", "comment_count": 14, "views": 0, "votes": 13 }, { "id": 200743, "title": "Private 58th solution", "comment_count": 5, "views": 0, "votes": 16 }, { "id": 200656, "title": "SVM/XGB on GPU (23 minutes training+inference time)", "comment_count": 2, "views": 0, "votes": 21 }, { "id": 200600, "title": "Silver to gold? All you need is augmentation", "comment_count": 9, "views": 0, "votes": 22 }, { "id": 200716, "title": "More Than A Competition !?", "comment_count": 1, "views": 0, "votes": 15 }, { "id": 201597, "title": "What is the innovatove technolgy used for building the dataset?", "comment_count": 1, "views": 0, "votes": null }, { "id": 200702, "title": "Mixup augmentaion on MoA tabular data", "comment_count": 7, "views": 0, "votes": 3 }, { "id": 200704, "title": "Our Bronze Medal Solution (if we had picked it)", "comment_count": 7, "views": 0, "votes": 12 }, { "id": 200642, "title": "First (serious) competition, first medal!!!", "comment_count": 5, "views": 0, "votes": 11 }, { "id": 200649, "title": "My 2 cents - Tabnet LB 0.01833 - Private 0.01625", "comment_count": 5, "views": 0, "votes": 14 }, { "id": 200034, "title": "Can't get better than 0.7, what I'm doing wrong?", "comment_count": 14, "views": 0, "votes": 2 }, { "id": 200087, "title": "Did you find any MoA MAGIC ?", "comment_count": 31, "views": 0, "votes": 17 }, { "id": 200528, "title": "Finally a competition master!", "comment_count": 10, "views": 0, "votes": 15 }, { "id": 200475, "title": "wow, new leaders of Public Leaderboard", "comment_count": 12, "views": 0, "votes": 10 }, { "id": 200577, "title": "The Shakeup visualization!", "comment_count": 11, "views": 0, "votes": 6 }, { "id": 200613, "title": "Private 55th place solution", "comment_count": 4, "views": 0, "votes": 11 }, { "id": 201160, "title": "CV unrelated to LB?", "comment_count": 1, "views": 0, "votes": 3 }, { "id": 201123, "title": "(Im?)possibility of (in?)appropriate use of domain data ?", "comment_count": 0, "views": 0, "votes": 4 }, { "id": 200551, "title": "TabNet.. TabNet... TabNet... How good was it?", "comment_count": 8, "views": 0, "votes": 8 }, { "id": 200569, "title": "lucky!!! 234th to 48th.share my solutions", "comment_count": 4, "views": 0, "votes": 13 }, { "id": 200984, "title": "NN Tuning Question", "comment_count": 1, "views": 0, "votes": 5 }, { "id": 200783, "title": "What environment did you use in this competition? ", "comment_count": 8, "views": 0, "votes": 3 }, { "id": 200675, "title": "Does folds number matter?", "comment_count": 7, "views": 0, "votes": 5 }, { "id": 200394, "title": "Anyone using data augmentation methods?", "comment_count": 19, "views": 0, "votes": 2 }, { "id": 199643, "title": "The sad truth about sklearn PCA", "comment_count": 12, "views": 0, "votes": 16 }, { "id": 200635, "title": "the most important points", "comment_count": 1, "views": 0, "votes": 6 }, { "id": 200529, "title": "Don't step out without scheduling : why the scheduler matter ?", "comment_count": 3, "views": 0, "votes": 11 }, { "id": 200607, "title": "Weight Optimization", "comment_count": 0, "views": 0, "votes": 10 }, { "id": 200526, "title": "Congrats to All!", "comment_count": 5, "views": 0, "votes": 6 }, { "id": 200603, "title": "Almost made the needed jump :(", "comment_count": 12, "views": 0, "votes": 6 }, { "id": 200325, "title": "Different CV and LB When Loading TabNet", "comment_count": 4, "views": 0, "votes": 4 }, { "id": 200590, "title": "56th solution and my share", "comment_count": 4, "views": 0, "votes": 7 }, { "id": 200541, "title": "20th place (If we had picked it)", "comment_count": 0, "views": 0, "votes": 10 }, { "id": 200396, "title": "Top 5 EDA notebooks : My list", "comment_count": 1, "views": 0, "votes": 12 }, { "id": 200572, "title": "51th solution", "comment_count": 3, "views": 0, "votes": 6 }, { "id": 198850, "title": "Advices for those tunning the tabnet model", "comment_count": 41, "views": 0, "votes": 61 }, { "id": 200524, "title": "Shocked by the shakeup", "comment_count": 11, "views": 0, "votes": 4 }, { "id": 200580, "title": "About Domain Knowledge", "comment_count": 3, "views": 0, "votes": 4 }, { "id": 200621, "title": "Stacked LSTM NN Starter code for Model Blending", "comment_count": 6, "views": 0, "votes": 3 }, { "id": 200503, "title": "Good luck to everyone ", "comment_count": 1, "views": 0, "votes": 6 }, { "id": 200386, "title": "Watch out for VarianceThreshold if you has independent inference notebook", "comment_count": 4, "views": 0, "votes": 6 }, { "id": 200678, "title": "Did your single class models perform well? ", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 200612, "title": "Anyone used KShot/Few Shot learning?", "comment_count": 2, "views": 0, "votes": 3 }, { "id": 200584, "title": "Making code snippets ", "comment_count": 7, "views": 0, "votes": null }, { "id": 200505, "title": "Just Curious Question about the timing", "comment_count": 2, "views": 0, "votes": 3 }, { "id": 200667, "title": "How are teams with the same score ranked?", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 200547, "title": "Drug Detection for MOA solution post processing", "comment_count": 0, "views": 0, "votes": 5 }, { "id": 199696, "title": "Hard Targets", "comment_count": 21, "views": 0, "votes": 15 }, { "id": 200576, "title": "R was a winner here (in a theoretical way)", "comment_count": 0, "views": 0, "votes": 4 }, { "id": 200417, "title": "Submission Timings", "comment_count": 3, "views": 0, "votes": 4 }, { "id": 200452, "title": "Soliciting Advice/Tips on NN/General Model Tuning Thread", "comment_count": 0, "views": 0, "votes": 5 }, { "id": 199959, "title": "Power of Groupby Methods", "comment_count": 3, "views": 0, "votes": 14 }, { "id": 200514, "title": "The submission I was going to choose just went doun with kaggle error", "comment_count": 6, "views": 0, "votes": null }, { "id": 200619, "title": "A question about private Leaderboard", "comment_count": 4, "views": 0, "votes": null }, { "id": 200565, "title": "Dear kaggle organizer（public37-private4000+）", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 200601, "title": "How to get medal with less 10 entries?", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 200583, "title": "How to blend model!!!I drop more than 4000!!!", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 200558, "title": "I think there is a problem with my score(public37 - private4000+?)", "comment_count": 19, "views": 0, "votes": -2 }, { "id": 199355, "title": "[Startup] One Click Deep Learning Platform with a $20 free credit on signup.", "comment_count": 14, "views": 0, "votes": 10 }, { "id": 197972, "title": "RAPIDS - Genetic Algorithm - Now Available in Kaggle Notebooks!", "comment_count": 24, "views": 0, "votes": 95 }, { "id": 197654, "title": "Lish MOA : (almost) winning ideas", "comment_count": 34, "views": 0, "votes": 64 }, { "id": 199484, "title": "Shake Up coming...?", "comment_count": 25, "views": 0, "votes": 15 }, { "id": 200655, "title": "what did u learn from this competition?", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 199516, "title": "My Hint to avoid Shake-up in this Competition", "comment_count": 6, "views": 0, "votes": 12 }, { "id": 200556, "title": "TabNet Tuning", "comment_count": 0, "views": 0, "votes": 2 }, { "id": 199983, "title": "Eat, sleep and Pray... for a limited shakeup!! ", "comment_count": 2, "views": 0, "votes": 6 }, { "id": 197914, "title": "Self-Supervised Pretraining with pytorch-tabnet", "comment_count": 7, "views": 0, "votes": 60 }, { "id": 200602, "title": "Single model CV 1559 | Public LB 1832 | Private LB 1624|No Blending", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 200368, "title": "Submission CSVs are identical up to 7 decimal places except for 2 cells out of 824,274 cells but the Public Score drops from 0.01830 to 0.13073", "comment_count": 3, "views": 0, "votes": null }, { "id": 199752, "title": " KISS Nuggets-01: Time your code in a simple way", "comment_count": 4, "views": 0, "votes": 5 }, { "id": 198898, "title": "Feature Engineering Ideas", "comment_count": 17, "views": 0, "votes": 40 }, { "id": 200711, "title": "Moa vs No MoA sample classification ", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 200360, "title": "How to format submission.csv!!", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 199303, "title": "Is data really randomly splitted?", "comment_count": 8, "views": 0, "votes": 22 }, { "id": 200589, "title": "Questions about the Variance Threshold Method", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 195788, "title": "QuantileTransformer parameters tuning", "comment_count": 9, "views": 0, "votes": 23 }, { "id": 200550, "title": "Any insight from People that were able to successfully use Non-Scored Targets? ", "comment_count": 1, "views": 0, "votes": null }, { "id": 199792, "title": "Notebook error after 3 minutes : Submission CSV Not Found", "comment_count": 9, "views": 0, "votes": 3 }, { "id": 199582, "title": "On submission, our code is run against 100 % of the test data (private+public) right ?", "comment_count": 10, "views": 0, "votes": 2 }, { "id": 199437, "title": "Some thoughts on the Contest", "comment_count": 3, "views": 0, "votes": 16 }, { "id": 200465, "title": "Is there anyone add test control to the training set?", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 198265, "title": "TabNet : Just for Curiosity !!!", "comment_count": 21, "views": 0, "votes": 1 }, { "id": 200567, "title": "Manual Submissions disabled", "comment_count": 1, "views": 0, "votes": -2 }, { "id": 199718, "title": "Error when submitting results ", "comment_count": 14, "views": 0, "votes": 0 }, { "id": 200349, "title": "The number of layers approximation", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 200214, "title": "Tabular Convolution", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 198088, "title": "Solution to the inconsistent scores with pretrained models", "comment_count": 21, "views": 0, "votes": 18 }, { "id": 198791, "title": "What to do in last week (final submission & shake-up)", "comment_count": 18, "views": 0, "votes": 8 }, { "id": 200121, "title": "Loading Trained Models as Datasets and Predicting on Inference Kernel", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 198970, "title": "Stochastic Weight Averaging", "comment_count": 7, "views": 0, "votes": 13 }, { "id": 200058, "title": "Re-run time 2 hours for GPU ?", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 199508, "title": "Maybe one smoothing does not fit all ...", "comment_count": 2, "views": 0, "votes": 8 }, { "id": 199248, "title": "Has Anyone tried Siamese Network?", "comment_count": 10, "views": 0, "votes": 1 }, { "id": 200135, "title": "Inference function for Tensorflow ", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 199390, "title": "Time limit", "comment_count": 13, "views": 0, "votes": 2 }, { "id": 199655, "title": "Questions about the Pre-Trained Model and Final Shake-up", "comment_count": 7, "views": 0, "votes": 2 }, { "id": 197361, "title": "Can I blend Python-based and R-based models in my kaggle kernel?", "comment_count": 11, "views": 0, "votes": 5 }, { "id": 199593, "title": "Same notebook differred in LB (Local inference LB:0183X vs kaggle kernel Inference: 0186X)", "comment_count": 5, "views": 0, "votes": 2 }, { "id": 199467, "title": "My top 20 of targets to consider for training mono-target model", "comment_count": 0, "views": 0, "votes": 6 }, { "id": 198724, "title": "transfer Learning with TabNet the probelm of final_mapping", "comment_count": 10, "views": 0, "votes": 10 }, { "id": 197108, "title": "The CV prediction for those aiming for money", "comment_count": 25, "views": 0, "votes": 35 }, { "id": 199304, "title": "Mega Issues with Submissions?", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 199123, "title": "KFold vs model.fit()", "comment_count": 5, "views": 0, "votes": 2 }, { "id": 198883, "title": "Question on pre-processing for inference part", "comment_count": 6, "views": 0, "votes": 8 }, { "id": 199211, "title": "A bit weird way of using VarianceThreshold", "comment_count": 3, "views": 0, "votes": 4 }, { "id": 198993, "title": "Sort by Best Score Not Working for Notebooks", "comment_count": 4, "views": 0, "votes": 3 }, { "id": 199584, "title": "ValueError: Expected more than 1 value per channel when training", "comment_count": 2, "views": 0, "votes": null }, { "id": 196625, "title": "how to load tabnet pre-trained models?", "comment_count": 21, "views": 0, "votes": 12 }, { "id": 193907, "title": "Best single model - 2nd month", "comment_count": 94, "views": 0, "votes": 82 }, { "id": 199682, "title": "help with submission error, please!", "comment_count": 4, "views": 0, "votes": null }, { "id": 199901, "title": "can any person explain about final model?", "comment_count": 2, "views": 0, "votes": -7 }, { "id": 199191, "title": "Anyone had recent submission issues?", "comment_count": 6, "views": 0, "votes": 1 }, { "id": 198246, "title": "Submission CSV Format Query!!!", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 199065, "title": "I got the same submission score for training and inference.", "comment_count": 0, "views": 0, "votes": 3 }, { "id": 198568, "title": "Help - how to load a tabnet model after saving?", "comment_count": 8, "views": 0, "votes": 1 }, { "id": 197962, "title": "Getting LB in .1820s ", "comment_count": 16, "views": 0, "votes": 10 }, { "id": 194345, "title": "FE for Pytorch-RankGauss-PCA-NN model with LB=0.01839", "comment_count": 7, "views": 0, "votes": 23 }, { "id": 198168, "title": "Early Stopping is one of the key differentiators in this Competition", "comment_count": 11, "views": 0, "votes": 11 }, { "id": 199041, "title": "Kaggle submission error", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 199430, "title": "Kaggle Notebooks Network Error?", "comment_count": 1, "views": 0, "votes": 0 }, { "id": 198577, "title": "PCA (and Variance Thresholding) for both train and test for preprocessing?", "comment_count": 12, "views": 0, "votes": 3 }, { "id": 199050, "title": "Kaggle system bugs?", "comment_count": 1, "views": 0, "votes": null }, { "id": 198364, "title": "What do you mean by 'blending' ?", "comment_count": 4, "views": 0, "votes": 7 }, { "id": 198706, "title": "Should I use this model for final ensemble?", "comment_count": 7, "views": 0, "votes": 2 }, { "id": 198298, "title": "Is This Overfitting?", "comment_count": 11, "views": 0, "votes": 4 }, { "id": 198964, "title": "New Way to Initialize the FC layer in Keras", "comment_count": 0, "views": 0, "votes": 2 }, { "id": 196830, "title": "Public TabNet model does not support pretraining", "comment_count": 9, "views": 0, "votes": 38 }, { "id": 197855, "title": "It is not only a multi-target problem", "comment_count": 8, "views": 0, "votes": 27 }, { "id": 193532, "title": "Be careful to shakeup.", "comment_count": 29, "views": 0, "votes": 77 }, { "id": 199147, "title": "Multiple models vs. a single model?", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 198509, "title": "Using unsupervised pretraining ", "comment_count": 6, "views": 0, "votes": 2 }, { "id": 198646, "title": "Submission CSV Not Found", "comment_count": 5, "views": 0, "votes": null }, { "id": 199116, "title": "Did anyone try to use GAN?", "comment_count": 1, "views": 0, "votes": null }, { "id": 197850, "title": "My model got overfit because...", "comment_count": 17, "views": 0, "votes": 13 }, { "id": 198401, "title": "Does weighting average overfit to OOF?", "comment_count": 8, "views": 0, "votes": 2 }, { "id": 198841, "title": "Finding Optimal Number of Components For Gene and Cell Features !", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 199055, "title": "Submission", "comment_count": 1, "views": 0, "votes": null }, { "id": 198902, "title": "How to make a final submission?", "comment_count": 3, "views": 0, "votes": null }, { "id": 198385, "title": "Mean of Logits vs Mean of Probabilities", "comment_count": 2, "views": 0, "votes": 3 }, { "id": 198531, "title": "Submission CSV Not Found", "comment_count": 2, "views": 0, "votes": null }, { "id": 199005, "title": "【Use TabNet for single-machine multi-graphics training】", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 198723, "title": "LB score of 0.020ish with a Neural Net model. Looking for a team member to share and get into top scores of 0.018ish..", "comment_count": 4, "views": 0, "votes": -9 }, { "id": 198775, "title": "Tensorflow: Differences in val scores between fit, evaluate and predict", "comment_count": 1, "views": 0, "votes": null }, { "id": 198633, "title": "Should we apply Label Smoothing during predictions?", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 198887, "title": "Time eror witр a ыги that takes less time", "comment_count": 2, "views": 0, "votes": null }, { "id": 197435, "title": "Tips for underfitting", "comment_count": 23, "views": 0, "votes": 13 }, { "id": 197650, "title": "[Lgbm Custom Loss] One Lgb for 206 Labels", "comment_count": 7, "views": 0, "votes": 22 }, { "id": 198289, "title": "Dose any one try to use GNN in this multilabel classification task", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 197250, "title": "Is this a proper implementation of Confidence Penalty?", "comment_count": 7, "views": 0, "votes": 17 }, { "id": 198548, "title": "About Run-time limit ?", "comment_count": 3, "views": 0, "votes": null }, { "id": 198220, "title": "How's the private leader board will be calculated?", "comment_count": 7, "views": 0, "votes": null }, { "id": 197376, "title": "a big difference of submission score between a full running notebook and a notebook with pre-trained model", "comment_count": 22, "views": 0, "votes": 11 }, { "id": 197077, "title": "Chris's Fold + Separated Inputs Ensemble", "comment_count": 10, "views": 0, "votes": 21 }, { "id": 197278, "title": "Output bias initialization", "comment_count": 20, "views": 0, "votes": 25 }, { "id": 197126, "title": "Two-Stage Learning of XGBoost", "comment_count": 12, "views": 0, "votes": 26 }, { "id": 197895, "title": "How to use \"BLEND\" in private LB?", "comment_count": 6, "views": 0, "votes": 4 }, { "id": 195932, "title": "(0.01835LB) Single model by transfer learning with non-scored label ", "comment_count": 31, "views": 0, "votes": 33 }, { "id": 198493, "title": "Merger deadline", "comment_count": 3, "views": 0, "votes": null }, { "id": 198461, "title": "Notebooks with gpu works more than 2 days", "comment_count": 2, "views": 0, "votes": null }, { "id": 197448, "title": "why NN when there is a less data?", "comment_count": 8, "views": 0, "votes": 5 }, { "id": 198253, "title": "PCA produces different results while run on different machines", "comment_count": 4, "views": 0, "votes": null }, { "id": 195642, "title": "Denoising auto-encoder as seen in Porto seguro's top solution", "comment_count": 17, "views": 0, "votes": 59 }, { "id": 197936, "title": " Clearing Blending confusions", "comment_count": 2, "views": 0, "votes": 4 }, { "id": 195368, "title": "Finally got Tabnet beat my other models with LB 0.01838", "comment_count": 24, "views": 0, "votes": 34 }, { "id": 197510, "title": "An Idea to Break the Degeneracy in CV/LB Relationship", "comment_count": 13, "views": 0, "votes": 8 }, { "id": 197728, "title": "Can't turn GPU on", "comment_count": 5, "views": 0, "votes": 5 }, { "id": 197661, "title": "I am getting constant LB score", "comment_count": 9, "views": 0, "votes": 1 }, { "id": 195888, "title": "CV Data Leakage Using PCA?", "comment_count": 10, "views": 0, "votes": 24 }, { "id": 198439, "title": "Can somebody help me out on how to submit my prediction file?", "comment_count": 1, "views": 0, "votes": null }, { "id": 198163, "title": "Did the \"Sort-By\" works well?", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 195660, "title": "Performance Degradation after Using the New Group CV", "comment_count": 32, "views": 0, "votes": 17 }, { "id": 198400, "title": "Thanks for sharing the infor!", "comment_count": 1, "views": 0, "votes": null }, { "id": 197951, "title": "Why \"Submit to competition\" has to run code again, as \"Save Version\" already run and generated submission.csv? ", "comment_count": 3, "views": 0, "votes": null }, { "id": 198015, "title": "LB score different from the public score", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 197475, "title": "What kind of ResNet are you using?", "comment_count": 4, "views": 0, "votes": 8 }, { "id": 192855, "title": "Single Model TabNet [LB - 0.01836][CV - 0.0166]", "comment_count": 46, "views": 0, "votes": 62 }, { "id": 197960, "title": "How to find correlation between different models?", "comment_count": 5, "views": 0, "votes": 0 }, { "id": 198233, "title": "Inserting the log_loss_metric in training loop ", "comment_count": 2, "views": 0, "votes": null }, { "id": 198234, "title": "Val AUC does not improve at all...", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 198211, "title": "My output is innverted??", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 197489, "title": "It is hard to break 0.018, but they did it.", "comment_count": 1, "views": 0, "votes": 3 }, { "id": 196320, "title": "Learning Rate Schedulers", "comment_count": 7, "views": 0, "votes": 18 }, { "id": 197425, "title": "Why my blending overfit ?", "comment_count": 7, "views": 0, "votes": 4 }, { "id": 197491, "title": "MoA - Interesting pattern in predictions ?", "comment_count": 0, "views": 0, "votes": 6 }, { "id": 192724, "title": "VarianceThreshold may cause \"Submission CSV Not Found\"", "comment_count": 8, "views": 0, "votes": 28 }, { "id": 197890, "title": "Why two same submission.csv with different public score ? ", "comment_count": 3, "views": 0, "votes": null }, { "id": 195859, "title": "Performance Influence of Pretraining on Non-Scored Targets", "comment_count": 24, "views": 0, "votes": 55 }, { "id": 197427, "title": "When blend - use the whole oof or split into folds and control on \"unseen\" part ?", "comment_count": 6, "views": 0, "votes": 4 }, { "id": 192932, "title": "New promising optimizer: AdaBelief", "comment_count": 9, "views": 0, "votes": 38 }, { "id": 196874, "title": "How do we end up with overfit to public LB?", "comment_count": 15, "views": 0, "votes": 17 }, { "id": 196436, "title": "GrowNet: Gradient Boosting Neural Networks", "comment_count": 5, "views": 0, "votes": 36 }, { "id": 197727, "title": "the code takes a very long time", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 197876, "title": "gpu running slower than cpu ?", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 197218, "title": "Label smoothing with TabNet Pytorch", "comment_count": 4, "views": 0, "votes": 3 }, { "id": 197391, "title": "Gauss rank conversion faster Calculation", "comment_count": 0, "views": 0, "votes": 4 }, { "id": 197588, "title": "Inference-which model to choose when have multi-models?", "comment_count": 4, "views": 0, "votes": null }, { "id": 197356, "title": "Combining Models Implementd with PyTorch and TF 2.x", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 197797, "title": "Find a teammate!", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 196703, "title": "Feature Importance in seconds with Rapids", "comment_count": 1, "views": 0, "votes": 33 }, { "id": 197158, "title": "Use num_threads = 4 for lightgbm on kaggle", "comment_count": 0, "views": 0, "votes": 8 }, { "id": 197213, "title": "how to keep prediction in submission file format?", "comment_count": 6, "views": 0, "votes": 2 }, { "id": 196463, "title": "Challenging points about Lish MOA", "comment_count": 5, "views": 0, "votes": 29 }, { "id": 197401, "title": "Difference in the value of train loss and val loss...", "comment_count": 1, "views": 0, "votes": 2 }, { "id": 197020, "title": "Any post process suggested?", "comment_count": 10, "views": 0, "votes": 4 }, { "id": 197296, "title": "MoA Mysterious Patterns for LocallyLinearEmbedding", "comment_count": 0, "views": 0, "votes": 3 }, { "id": 197191, "title": "Submission Scoring Error", "comment_count": 9, "views": 0, "votes": null }, { "id": 197315, "title": "What is the correct way to ensemble models?", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 194973, "title": "FE : VarianceThreshold - what else is there?", "comment_count": 22, "views": 0, "votes": 17 }, { "id": 197548, "title": "need help for my final project", "comment_count": 4, "views": 0, "votes": -16 }, { "id": 196958, "title": "\"SUBMISSION CSV NOT FOUND ERROR\" during [ Inference ] not while [Train+Inference] ?", "comment_count": 13, "views": 0, "votes": 1 }, { "id": 197284, "title": "Run kernel in two hour", "comment_count": 2, "views": 0, "votes": null }, { "id": 197472, "title": "Using Autoencoders", "comment_count": 1, "views": 0, "votes": null }, { "id": 197360, "title": "Are we provided the nonscored targets in test data? ", "comment_count": 3, "views": 0, "votes": null }, { "id": 195802, "title": "GAP between CV and LB", "comment_count": 4, "views": 0, "votes": 3 }, { "id": 196665, "title": "Question about competition rule \"including pre-trained models\"", "comment_count": 5, "views": 0, "votes": 7 }, { "id": 197132, "title": "Single Target Networks - Chris Splits ", "comment_count": 10, "views": 0, "votes": 1 }, { "id": 196081, "title": "Transformers are not Transformers", "comment_count": 13, "views": 0, "votes": 32 }, { "id": 197083, "title": "What is the random LB variation that can be expected?", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 197048, "title": "Trouble submitting - always getting error 'Submission Scoring Error'", "comment_count": 5, "views": 0, "votes": 2 }, { "id": 197390, "title": "Did TTA work for anyone?", "comment_count": 1, "views": 0, "votes": null }, { "id": 196504, "title": "Remark on CV vs LB misfit", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 196113, "title": "Don't increase your Folds anymore", "comment_count": 13, "views": 0, "votes": 11 }, { "id": 196099, "title": "How low will it get?", "comment_count": 8, "views": 0, "votes": 3 }, { "id": 197011, "title": "Does the test set include private set? ", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 197309, "title": "Submission error: Submission CSV Not Found", "comment_count": 1, "views": 0, "votes": null }, { "id": 194219, "title": "Any clues for 'Submission CSV not found' error", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 197241, "title": "Whether the public kernels have overfitting?", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 196279, "title": "Set as Utility Script", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 192406, "title": "\"Submission CSV Not Found\" I need help.", "comment_count": 3, "views": 0, "votes": 4 }, { "id": 196319, "title": "XGBoost multi-label training strategies?", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 196696, "title": "Ruined by \"zoombombers\" - beware. A zoom seminar talk on bio background for the MoA competition (in Russian, sorry )", "comment_count": 1, "views": 0, "votes": 3 }, { "id": 196232, "title": "Which model should I choose for my Final submission?", "comment_count": 9, "views": 0, "votes": 9 }, { "id": 196364, "title": "Identical Submissions and different scores", "comment_count": 9, "views": 0, "votes": 5 }, { "id": 196592, "title": "Does submit run on entire test set or only 25% for public leaderboard", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 196258, "title": "Gradient Surgery for Multi-Task Learning", "comment_count": 1, "views": 0, "votes": 14 }, { "id": 197205, "title": "Can´t create an R script notebook", "comment_count": 2, "views": 0, "votes": null }, { "id": 191545, "title": "Ideas to tackle class imbalance problem", "comment_count": 36, "views": 0, "votes": 88 }, { "id": 195928, "title": "Failed to optimise hyperparameters with Optuna using the new CV scheme", "comment_count": 15, "views": 0, "votes": 5 }, { "id": 196560, "title": "pdgfr_inhibitor and other targets well predicted by boostings ", "comment_count": 0, "views": 0, "votes": 6 }, { "id": 196524, "title": "Which CV score is better, average oof log loss or log loss with all oof predictions ?", "comment_count": 3, "views": 0, "votes": 3 }, { "id": 197066, "title": "Question with tuning a NN model with tuning_run", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 196952, "title": "mlsmote cuda compatible", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 196224, "title": "Questions about the Time Limit", "comment_count": 4, "views": 0, "votes": 7 }, { "id": 195884, "title": "Old CV + Drugs CV ensemble", "comment_count": 5, "views": 0, "votes": 3 }, { "id": 192410, "title": "How should we use Pseudo Labelling?", "comment_count": 17, "views": 0, "votes": 6 }, { "id": 195016, "title": "Be very careful about training and inferencing code", "comment_count": 5, "views": 0, "votes": 13 }, { "id": 196283, "title": "XGBoost MultiOutputClassifier saving?", "comment_count": 0, "views": 0, "votes": 6 }, { "id": 188635, "title": "Any ideas for data augmentation for under-represented classes?", "comment_count": 6, "views": 0, "votes": 8 }, { "id": 195810, "title": " Submission CSV not found error", "comment_count": 5, "views": 0, "votes": 3 }, { "id": 194203, "title": "What can TabNet see", "comment_count": 44, "views": 0, "votes": 31 }, { "id": 196726, "title": "What is the memory size available to us in CPU/GPU/TPU ?", "comment_count": 5, "views": 0, "votes": null }, { "id": 195895, "title": "Using scipy for blending weights optimisation with gradients ", "comment_count": 2, "views": 0, "votes": 19 }, { "id": 195882, "title": "Change in the way Notebooks are created ????", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 195891, "title": "High Dropout Ensemble Learning", "comment_count": 4, "views": 0, "votes": 6 }, { "id": 190968, "title": "Submission error and column \"X5.alpha_reductase_inhibitor\"", "comment_count": 3, "views": 0, "votes": null }, { "id": 196784, "title": "Can we train models outside Kaggle ?", "comment_count": 2, "views": 0, "votes": null }, { "id": 193984, "title": "Does Changing Dosage/time for a particular Drug affects the targets?", "comment_count": 22, "views": 0, "votes": 7 }, { "id": 195179, "title": "New file released! What does it mean?", "comment_count": 17, "views": 0, "votes": 39 }, { "id": 195706, "title": "Server error, cannot submit", "comment_count": 8, "views": 0, "votes": 9 }, { "id": 195731, "title": "Notes about TabNet in the public Notebook", "comment_count": 7, "views": 0, "votes": 9 }, { "id": 194976, "title": "CTGAN Data Augmentation", "comment_count": 19, "views": 0, "votes": 34 }, { "id": 193583, "title": "Learning Label Correlation by Sequential Models", "comment_count": 18, "views": 0, "votes": 40 }, { "id": 196413, "title": "[Beginner] Score gap between CV and OOF", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 193999, "title": "Beginner question: Train loss vs Valid Loss", "comment_count": 11, "views": 0, "votes": 5 }, { "id": 195969, "title": "Group of targets", "comment_count": 0, "views": 0, "votes": 5 }, { "id": 194422, "title": "Different calculation local/Kaggle platform", "comment_count": 12, "views": 0, "votes": 2 }, { "id": 196096, "title": "XGBoost Application", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 195760, "title": "Variance Thresholding and Submission Notebook", "comment_count": 5, "views": 0, "votes": 5 }, { "id": 196193, "title": "Error importing Tensorflow ", "comment_count": 1, "views": 0, "votes": 2 }, { "id": 195585, "title": "Random Ensemble Neural Network", "comment_count": 2, "views": 0, "votes": 10 }, { "id": 193708, "title": "Why it might be difficult to break 0.01800?", "comment_count": 13, "views": 0, "votes": 29 }, { "id": 196006, "title": "Gene expression formula", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 185161, "title": "A method to smart-initialize the final layers of NNs", "comment_count": 34, "views": 0, "votes": 56 }, { "id": 195483, "title": "Using keywords to create new targets", "comment_count": 5, "views": 0, "votes": 11 }, { "id": 194080, "title": "Dosage and Time Affects the Targets", "comment_count": 7, "views": 0, "votes": 4 }, { "id": 195217, "title": "Extracting insights from drug_id", "comment_count": 2, "views": 0, "votes": 48 }, { "id": 195227, "title": "What's your CV-LB with Drug and MultiLabel Stratification?", "comment_count": 10, "views": 0, "votes": 10 }, { "id": 195380, "title": "Transfer learning idea on the NEW file (Using new file as feature)", "comment_count": 3, "views": 0, "votes": 21 }, { "id": 188452, "title": "Best stacking library?", "comment_count": 6, "views": 0, "votes": 17 }, { "id": 188651, "title": "Neural Network train & val loss blow up suddenly ", "comment_count": 26, "views": 0, "votes": 31 }, { "id": 196116, "title": "GPU quote set to 0?", "comment_count": 3, "views": 0, "votes": null }, { "id": 195544, "title": "Simple stats on the top 8 drug_ids", "comment_count": 7, "views": 0, "votes": 8 }, { "id": 194190, "title": "Revealing the drug IDs", "comment_count": 43, "views": 0, "votes": 44 }, { "id": 195472, "title": "Insights in Cp_time and Cp_dose for finding effective features ?", "comment_count": 2, "views": 0, "votes": 8 }, { "id": 191691, "title": "Are we close to confirming NN fits best in this competition?", "comment_count": 36, "views": 0, "votes": 42 }, { "id": 193483, "title": "Would this help in Blending?", "comment_count": 15, "views": 0, "votes": 3 }, { "id": 196060, "title": "Where is my output file ?", "comment_count": 2, "views": 0, "votes": null }, { "id": 195546, "title": "Why not implement Rank Gauss to all feature columns.", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 195751, "title": "Final_Inference Kernel", "comment_count": 8, "views": 0, "votes": 2 }, { "id": 193363, "title": "Silly question probably.", "comment_count": 36, "views": 0, "votes": 29 }, { "id": 195561, "title": "Average CV predictions or re-train the full dataset?", "comment_count": 5, "views": 0, "votes": 3 }, { "id": 196037, "title": "How to submitte my csv file as a result", "comment_count": 1, "views": 0, "votes": null }, { "id": 195907, "title": "Must we train the model within the 9 hours (CPU) or 2 hours (GPU) limit?", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 195764, "title": "Representativeness of train data", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 195243, "title": "T Test (Student’s T-Test) can be used for Feature Selection", "comment_count": 2, "views": 0, "votes": 7 }, { "id": 193918, "title": "I'm quite confused and stress about next a month", "comment_count": 19, "views": 0, "votes": 9 }, { "id": 195539, "title": "why score not same?", "comment_count": 8, "views": 0, "votes": 0 }, { "id": 190838, "title": "A Tip on Parameter Tuning for Beginners", "comment_count": 9, "views": 0, "votes": 12 }, { "id": 195339, "title": "A lecture on competition via zoom (in Russian, sorry)", "comment_count": 0, "views": 0, "votes": 6 }, { "id": 195702, "title": "Learning resources for folks interested in this domain", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 195797, "title": "How are train_targets_scored and train_features related? ", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 194968, "title": "Iterative-Stratification: Seed Selection", "comment_count": 1, "views": 0, "votes": 14 }, { "id": 195289, "title": "Leaderboard ", "comment_count": 5, "views": 0, "votes": -1 }, { "id": 195366, "title": "How is MoA recorded?", "comment_count": 1, "views": 0, "votes": 2 }, { "id": 194847, "title": "Improving score by using less data?", "comment_count": 2, "views": 0, "votes": 12 }, { "id": 195382, "title": "Blending___", "comment_count": 5, "views": 0, "votes": 2 }, { "id": 194744, "title": "[SAVE WEEKLY GPU QUOTA/TIME DURING COMMIT]", "comment_count": 0, "views": 0, "votes": 12 }, { "id": 193985, "title": "Label Smoothing not working on TabNet", "comment_count": 5, "views": 0, "votes": 8 }, { "id": 194464, "title": "Has anyone used Captum or other feature importance tool with PyTorch model?", "comment_count": 3, "views": 0, "votes": 9 }, { "id": 193684, "title": "(L.B.- 0.01838 C.V.- 0.01797) Single Model & Understanding", "comment_count": 27, "views": 0, "votes": 62 }, { "id": 190163, "title": "How do you analyze model's performance?", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 189508, "title": "Important info: How to structure your submission notebook", "comment_count": 43, "views": 0, "votes": 71 }, { "id": 193586, "title": "(Denoising) Autoencoder", "comment_count": 5, "views": 0, "votes": 7 }, { "id": 192686, "title": "Query on Blending", "comment_count": 7, "views": 0, "votes": 7 }, { "id": 194374, "title": "Keras vs Pytorch Applying Features Engineering. ", "comment_count": 10, "views": 0, "votes": 5 }, { "id": 187946, "title": "A list of Github repos to use for this competition", "comment_count": 30, "views": 0, "votes": 118 }, { "id": 195092, "title": "SELU as activation function", "comment_count": 5, "views": 0, "votes": 2 }, { "id": 195349, "title": "Are the drug Ids enough? Don't we need a tissue id?", "comment_count": 1, "views": 0, "votes": null }, { "id": 195023, "title": "new to code competition", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 193987, "title": "Submission CSV not Found Error", "comment_count": 5, "views": 0, "votes": null }, { "id": 194182, "title": "skew, kurt or median as features ", "comment_count": 3, "views": 0, "votes": 10 }, { "id": 193909, "title": "what is your working environment?", "comment_count": 10, "views": 0, "votes": 5 }, { "id": 195136, "title": "Question: Why does LB performance decrease after FE?", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 193702, "title": "An explanation for why cross validation & leader board scores are different and what to do about it ", "comment_count": 15, "views": 0, "votes": 39 }, { "id": 194572, "title": "Offline lightgbm + GPU", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 193717, "title": "Quantile Normalization Paper", "comment_count": 11, "views": 0, "votes": 19 }, { "id": 191857, "title": "Strange instance in the dataset", "comment_count": 22, "views": 0, "votes": 34 }, { "id": 187212, "title": "Feature Engineering? How many features do you use?", "comment_count": 27, "views": 0, "votes": 22 }, { "id": 190070, "title": "Submit Predictions's doubts", "comment_count": 4, "views": 0, "votes": null }, { "id": 194050, "title": "Reason for 6 times co-occurrences?", "comment_count": 4, "views": 0, "votes": 4 }, { "id": 193878, "title": "magic again", "comment_count": 12, "views": 0, "votes": 9 }, { "id": 194161, "title": "Offline training, online reasoning integration", "comment_count": 11, "views": 0, "votes": 1 }, { "id": 181040, "title": "novel (?) ideas", "comment_count": 70, "views": 0, "votes": 154 }, { "id": 188978, "title": "RAPIDS is awesome!", "comment_count": 3, "views": 0, "votes": 43 }, { "id": 194302, "title": "Moa scored targets : Stats", "comment_count": 0, "views": 0, "votes": 3 }, { "id": 194956, "title": "Blind or not-blind cv (cross validation) score? ", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 193953, "title": "Fold vs Seed, the numbers", "comment_count": 3, "views": 0, "votes": 9 }, { "id": 193380, "title": "Same submission scored 0.01838 and 0.01843", "comment_count": 12, "views": 0, "votes": 12 }, { "id": 193329, "title": "LGBMClassifier with optimal param for each target", "comment_count": 10, "views": 0, "votes": 12 }, { "id": 193892, "title": "Label smoothing, CV and LB", "comment_count": 13, "views": 0, "votes": 2 }, { "id": 182818, "title": "Submission Issue", "comment_count": 3, "views": 0, "votes": null }, { "id": 194351, "title": "Error occurs when uploading trained models to submit", "comment_count": 3, "views": 0, "votes": null }, { "id": 191037, "title": "Different public scores between [train+inference] and the standalone [inference] kernels", "comment_count": 14, "views": 0, "votes": 3 }, { "id": 193463, "title": "CV (0.01460) L.B (0.01838) Single NN Model", "comment_count": 1, "views": 0, "votes": 9 }, { "id": 188860, "title": "Training & Inference", "comment_count": 13, "views": 0, "votes": 13 }, { "id": 187516, "title": "Doubt about Non Scored Targets", "comment_count": 15, "views": 0, "votes": 8 }, { "id": 193866, "title": "Noob question: pytorch multi-label classification", "comment_count": 15, "views": 0, "votes": 1 }, { "id": 193507, "title": "Different scores while inference", "comment_count": 9, "views": 0, "votes": 3 }, { "id": 194270, "title": "submission.csv not visible", "comment_count": 1, "views": 0, "votes": 0 }, { "id": 193376, "title": "Pytorch vs Tensorflow/Keras Performance", "comment_count": 21, "views": 0, "votes": 10 }, { "id": 188203, "title": "Feature Creation via Genetic Programming", "comment_count": 5, "views": 0, "votes": 29 }, { "id": 193239, "title": "Proper post-processing can give some improvement ", "comment_count": 15, "views": 0, "votes": 16 }, { "id": 193139, "title": "CUDA error: device-side assert triggered", "comment_count": 22, "views": 0, "votes": 4 }, { "id": 193358, "title": "Synthetic data and class imbalance", "comment_count": 9, "views": 0, "votes": 12 }, { "id": 190481, "title": "End2end implementation using Autoencoder", "comment_count": 7, "views": 0, "votes": 10 }, { "id": 181690, "title": "Has anyone tried TabNet in this competition?", "comment_count": 21, "views": 0, "votes": 12 }, { "id": 190128, "title": "help, I cannot submit : ModuleNotFoundError: No module named 'iterstrat'", "comment_count": 11, "views": 0, "votes": 7 }, { "id": 180304, "title": "About \"Controlled Perturbations\"", "comment_count": 37, "views": 0, "votes": 53 }, { "id": 193045, "title": "L.B. Vs C.V. Correlation (of My single model so far )", "comment_count": 20, "views": 0, "votes": 5 }, { "id": 193744, "title": "Post-processing with cp_type = ctr_vehicule", "comment_count": 3, "views": 0, "votes": null }, { "id": 187208, "title": "Submission CSV Not Found", "comment_count": 14, "views": 0, "votes": null }, { "id": 193155, "title": "output_files", "comment_count": 2, "views": 0, "votes": 3 }, { "id": 193493, "title": "Doubt regarding active competitions..", "comment_count": 5, "views": 0, "votes": 1 }, { "id": 192980, "title": "Can you enjoy competing 0.0001 by only blending?", "comment_count": 14, "views": 0, "votes": 6 }, { "id": 193662, "title": "xavier_normal ?", "comment_count": 6, "views": 0, "votes": -2 }, { "id": 192491, "title": "Anomaly Detection Approach", "comment_count": 4, "views": 0, "votes": 11 }, { "id": 193666, "title": "Drop cp_type?", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 193571, "title": "Logloss vs roc auc", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 193129, "title": "Tackling any Kaggle competition : The Noob's way", "comment_count": 1, "views": 0, "votes": 7 }, { "id": 193099, "title": "Seed Averaging", "comment_count": 10, "views": 0, "votes": 3 }, { "id": 193882, "title": "I think erbb2_inhibitor shouldn't count on the final score", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 193729, "title": "Show Distribution of Data", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 193646, "title": "Question about private inference submission blending", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 185096, "title": "How to compute class weights for multi-label problem?", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 188170, "title": "Metric for measuring performance on minority class?", "comment_count": 2, "views": 0, "votes": 7 }, { "id": 193298, "title": "What next to go beyond 0.02648?", "comment_count": 6, "views": 0, "votes": 2 }, { "id": 191487, "title": "Couple more questions on the data...", "comment_count": 10, "views": 0, "votes": 32 }, { "id": 192440, "title": "CV vs label smoothing", "comment_count": 31, "views": 0, "votes": 17 }, { "id": 193611, "title": "about \"the re-run dataset has approximately 4x the number of examples seen in the Public test.\"", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 192780, "title": " L.B.-(0.01839 --> 0.01837) Single NN Model - 1 Tip ", "comment_count": 20, "views": 0, "votes": 14 }, { "id": 193683, "title": "submission error", "comment_count": 1, "views": 0, "votes": null }, { "id": 192948, "title": "What's the best Scheduler?", "comment_count": 11, "views": 0, "votes": 5 }, { "id": 193498, "title": "Are provided data(g-/c-) already normalized? Could we transform to raw data? ", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 191135, "title": "Adjusting most imbalanced targets improves score", "comment_count": 27, "views": 0, "votes": 31 }, { "id": 193213, "title": "How can we know the CV is \"robust\"?", "comment_count": 6, "views": 0, "votes": null }, { "id": 193413, "title": "Model fusion", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 191578, "title": "6 experiments on a drug divided into train and test?", "comment_count": 9, "views": 0, "votes": 6 }, { "id": 193037, "title": "I am a new kaggler, if you don't mind, I would like to ask you a question.", "comment_count": 6, "views": 0, "votes": 3 }, { "id": 193492, "title": "Torch Sampler", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 193473, "title": "NON scored targets", "comment_count": 3, "views": 0, "votes": null }, { "id": 192785, "title": "Understanding of Label Smoothing", "comment_count": 3, "views": 0, "votes": 6 }, { "id": 193112, "title": "Submission Error", "comment_count": 1, "views": 0, "votes": null }, { "id": 192525, "title": "2 Newbie questions about train/test drug split and private vs public test data", "comment_count": 4, "views": 0, "votes": 11 }, { "id": 193176, "title": "How can i deal with this error?help please", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 192947, "title": "clip in labels smoothing", "comment_count": 2, "views": 0, "votes": null }, { "id": 192714, "title": "Reproduce clustering of MoA in umap projection obtained in recommended paper Nature Cancer, 2020", "comment_count": 3, "views": 0, "votes": 6 }, { "id": 191969, "title": "What happened to @blindape's post?", "comment_count": 10, "views": 0, "votes": 21 }, { "id": 192733, "title": "Questions about submission results", "comment_count": 2, "views": 0, "votes": 5 }, { "id": 192715, "title": " Is it normal to blend your results with others? What is the best single model score? ", "comment_count": 8, "views": 0, "votes": 3 }, { "id": 193115, "title": "Submission Error", "comment_count": 8, "views": 0, "votes": -6 }, { "id": 187288, "title": "CV vs LB ( Have no idea what's going on ....)", "comment_count": 33, "views": 0, "votes": 25 }, { "id": 192329, "title": "Would the model perform better if...?", "comment_count": 8, "views": 0, "votes": 7 }, { "id": 183032, "title": "Does the evaluation use the GPU?", "comment_count": 6, "views": 0, "votes": null }, { "id": 188624, "title": "Best Single model - 1st month", "comment_count": 83, "views": 0, "votes": 17 }, { "id": 192102, "title": "Question about label smoothing for binary cross entropy and logloss evaluation metric", "comment_count": 8, "views": 0, "votes": null }, { "id": 182002, "title": "Results worse with focal loss", "comment_count": 5, "views": 0, "votes": 4 }, { "id": 192244, "title": "using test_features for training?", "comment_count": 3, "views": 0, "votes": 5 }, { "id": 192537, "title": "Notebook running time when scoring private test set", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 192564, "title": "Questions about blending", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 192183, "title": "Are multiple notebooks allowed ?", "comment_count": 3, "views": 0, "votes": 8 }, { "id": 190808, "title": "How were training and test data split?", "comment_count": 3, "views": 0, "votes": 12 }, { "id": 183333, "title": "how to convert model.predict_proba into the submission form?", "comment_count": 2, "views": 0, "votes": null }, { "id": 192074, "title": "Motivation behind separating data by dose/time", "comment_count": 0, "views": 0, "votes": 10 }, { "id": 192798, "title": "Feature pre-processing only once or by fold?", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 191706, "title": "Agonist and Antagonist", "comment_count": 11, "views": 0, "votes": 15 }, { "id": 191830, "title": "Does Label Smoothing make LB-CV correlation worse?", "comment_count": 5, "views": 0, "votes": 1 }, { "id": 191848, "title": "Create a Control Group and Remove Noise", "comment_count": 2, "views": 0, "votes": 7 }, { "id": 181965, "title": "Looking for Common Causes of \"Submission Scoring Error\"", "comment_count": 11, "views": 0, "votes": 3 }, { "id": 191907, "title": "overfit public LB", "comment_count": 4, "views": 0, "votes": 4 }, { "id": 191079, "title": "Robust CV strategy?", "comment_count": 14, "views": 0, "votes": 14 }, { "id": 192482, "title": "Need help with submission error", "comment_count": 6, "views": 0, "votes": null }, { "id": 185593, "title": "After label smoothing, my CV matches my LB. Is this a coincidence?", "comment_count": 33, "views": 0, "votes": 28 }, { "id": 191818, "title": " Question about Blend(Weighted Average)", "comment_count": 4, "views": 0, "votes": 3 }, { "id": 183010, "title": "multi label log loss ", "comment_count": 8, "views": 0, "votes": 1 }, { "id": 191097, "title": "Pytorch-tabnet version 2.0.1 dataset", "comment_count": 11, "views": 0, "votes": 16 }, { "id": 191392, "title": "What is the utility of blending? ", "comment_count": 13, "views": 0, "votes": 6 }, { "id": 190194, "title": "Submission problem", "comment_count": 4, "views": 0, "votes": 3 }, { "id": 191267, "title": "de-anonymizing gene features", "comment_count": 6, "views": 0, "votes": 5 }, { "id": 191305, "title": "Neural Network swish Hidden Layer Activation", "comment_count": 12, "views": 0, "votes": 6 }, { "id": 191939, "title": "Is there a way I can upload my inference result in to Kaggle/working?", "comment_count": 3, "views": 0, "votes": null }, { "id": 191456, "title": "Simple TabNetRegressor", "comment_count": 0, "views": 0, "votes": 10 }, { "id": 190949, "title": "domain difference between train and public test", "comment_count": 6, "views": 0, "votes": 26 }, { "id": 191280, "title": "Offline Autograd Library for Blending", "comment_count": 5, "views": 0, "votes": 8 }, { "id": 190998, "title": "How to evaluate cell viability values?", "comment_count": 9, "views": 0, "votes": 20 }, { "id": 191620, "title": "Polypharmacy and This Competition", "comment_count": 2, "views": 0, "votes": 4 }, { "id": 191670, "title": "Unable to Train NODE on TPU", "comment_count": 5, "views": 0, "votes": 2 }, { "id": 187419, "title": "Upsampling Multilabeled Data with MLSMOTE", "comment_count": 32, "views": 0, "votes": 38 }, { "id": 191099, "title": "Submitting Submission.csv without computations", "comment_count": 7, "views": 0, "votes": 5 }, { "id": 191210, "title": "How to deal with Notebook Timeout Error", "comment_count": 7, "views": 0, "votes": 4 }, { "id": 190693, "title": "Experimental methods of this competition", "comment_count": 2, "views": 0, "votes": 33 }, { "id": 191667, "title": "Using pytorch-geometric!", "comment_count": 2, "views": 0, "votes": null }, { "id": 192129, "title": "Team_______", "comment_count": 0, "views": 0, "votes": -8 }, { "id": 189857, "title": "Did anyone else observe worse LB and CV after label smoothing?", "comment_count": 21, "views": 0, "votes": 24 }, { "id": 191409, "title": "How is the test data divided?", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 185650, "title": "Model Blending Possibility", "comment_count": 8, "views": 0, "votes": 49 }, { "id": 191254, "title": "Organize posts that help me", "comment_count": 1, "views": 0, "votes": 4 }, { "id": 191406, "title": "Different results from submission kernels", "comment_count": 1, "views": 0, "votes": 0 }, { "id": 180578, "title": "Save notebook commit time with these lines of code", "comment_count": 6, "views": 0, "votes": 5 }, { "id": 182387, "title": "Should Kaggle Limit publication of high scoring public kernels? ", "comment_count": 51, "views": 0, "votes": 24 }, { "id": 190708, "title": "magic number 0.3372", "comment_count": 5, "views": 0, "votes": 0 }, { "id": 182183, "title": "FYI: I trained a model. What is next?", "comment_count": 12, "views": 0, "votes": 45 }, { "id": 190428, "title": "What's the logic behind counting error submissions as valid submissions?", "comment_count": 9, "views": 0, "votes": 8 }, { "id": 189595, "title": "Are Shallow NN performing Better?", "comment_count": 5, "views": 0, "votes": 10 }, { "id": 190010, "title": "Importance of Sampling", "comment_count": 9, "views": 0, "votes": 4 }, { "id": 190684, "title": "Submission Scoring Error (Help)", "comment_count": 10, "views": 0, "votes": 3 }, { "id": 190655, "title": "Which algorithms can handle multiclassification problems ?", "comment_count": 2, "views": 0, "votes": 3 }, { "id": 189312, "title": "Is there any way to make clusters by drug names?", "comment_count": 11, "views": 0, "votes": 7 }, { "id": 190438, "title": "WHY NN ARE SUCCEDING", "comment_count": 2, "views": 0, "votes": 3 }, { "id": 180392, "title": "Looking to either join or form a team! ", "comment_count": 3, "views": 0, "votes": null }, { "id": 181164, "title": "which framework / nn architecture to use", "comment_count": 4, "views": 0, "votes": null }, { "id": 190199, "title": "Multilabel classification using deep learning?", "comment_count": 6, "views": 0, "votes": 6 }, { "id": 183003, "title": "Team Collaboration", "comment_count": 3, "views": 0, "votes": 0 }, { "id": 190230, "title": "Autoencoder reconstruction loss higher than PCA?", "comment_count": 2, "views": 0, "votes": 3 }, { "id": 190986, "title": "Can I Use Private Utility Scripts?", "comment_count": 1, "views": 0, "votes": null }, { "id": 191036, "title": "Submit error times will adding to all submit times?", "comment_count": 0, "views": 0, "votes": -2 }, { "id": 189368, "title": "Tracking Experiments with No Internet", "comment_count": 8, "views": 0, "votes": 2 }, { "id": 189135, "title": "Huge multicollinearity amongst features", "comment_count": 7, "views": 0, "votes": 3 }, { "id": 188710, "title": "Congrats to kxx for becoming Notebooks GM!", "comment_count": 12, "views": 0, "votes": 23 }, { "id": 180390, "title": "Gene expression interpretation", "comment_count": 4, "views": 0, "votes": 17 }, { "id": 190068, "title": "About TF 2.x and PyTorch", "comment_count": 5, "views": 0, "votes": 2 }, { "id": 188379, "title": "using models that was trained over 9 hours", "comment_count": 5, "views": 0, "votes": 2 }, { "id": 187357, "title": "Visual Guide to Datasets", "comment_count": 6, "views": 0, "votes": 21 }, { "id": 190381, "title": "Simple question about log_loss in this task", "comment_count": 10, "views": 0, "votes": null }, { "id": 190369, "title": "Estimating LB Score using existing submissions", "comment_count": 0, "views": 0, "votes": 2 }, { "id": 189290, "title": "from 0.01857 to 0.01886 ！！！Set samples' targets('atp-sensitive_potassium_channel_antagonist', 'erbb2_inhibitor') with [cp_dose = 'd2'] = 0 makes my lb worse", "comment_count": 5, "views": 0, "votes": 5 }, { "id": 182357, "title": "TypeError: A sparse matrix was passed, but dense data is required. Use X.toarray() to convert to a dense numpy array.", "comment_count": 4, "views": 0, "votes": 3 }, { "id": 185355, "title": "More single model possibility", "comment_count": 68, "views": 0, "votes": 110 }, { "id": 189678, "title": "Any approaches other than NN", "comment_count": 5, "views": 0, "votes": 3 }, { "id": 189053, "title": "Where to look when Validation improves, but LB doesn't?", "comment_count": 10, "views": 0, "votes": 10 }, { "id": 190399, "title": "Doubt on \"External Data and Pre-trained model\"", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 190044, "title": "how do you check new features?", "comment_count": 0, "views": 0, "votes": 4 }, { "id": 189932, "title": "Deal with highly imbalanced dataset", "comment_count": 1, "views": 0, "votes": 2 }, { "id": 190242, "title": "Problem - Model kept in memory even when deleted", "comment_count": 1, "views": 0, "votes": null }, { "id": 189934, "title": "Analyzing the clusters in the data", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 189146, "title": "Beginner understanding problem? ", "comment_count": 3, "views": 0, "votes": null }, { "id": 189680, "title": "How to submit the predictions", "comment_count": 5, "views": 0, "votes": null }, { "id": 187933, "title": "Multi Label Classification and Neural Networks", "comment_count": 7, "views": 0, "votes": 2 }, { "id": 186539, "title": "Blending Multilabeled Models with Scipy", "comment_count": 11, "views": 0, "votes": 9 }, { "id": 189345, "title": "Submission bug?", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 188761, "title": "Removing 'ctl_vehicle' from training data is worsening my CV. ", "comment_count": 12, "views": 0, "votes": 4 }, { "id": 183684, "title": "How to deal with train_targets_scored columns that have only one value for one label?", "comment_count": 10, "views": 0, "votes": 6 }, { "id": 189405, "title": "Submission Taking too long", "comment_count": 9, "views": 0, "votes": null }, { "id": 188525, "title": "Weight Normalization VS Batch Normalization", "comment_count": 7, "views": 0, "votes": 17 }, { "id": 180092, "title": "Multi-label competitions with tabular datasets", "comment_count": 6, "views": 0, "votes": 75 }, { "id": 189739, "title": "Evaluation precision in Tensorflow?", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 188814, "title": "How many models can I blend safely?", "comment_count": 10, "views": 0, "votes": 2 }, { "id": 187439, "title": "Experimental design of PRISM?, one gene expression and 100 cell viability.", "comment_count": 8, "views": 0, "votes": 5 }, { "id": 187121, "title": "Stuck at .0190", "comment_count": 18, "views": 0, "votes": 9 }, { "id": 188626, "title": "My submission is not showing up!", "comment_count": 30, "views": 0, "votes": -1 }, { "id": 188996, "title": "Confused beginner", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 187861, "title": "How to do Feature Creation - Brute Force ?", "comment_count": 6, "views": 0, "votes": 9 }, { "id": 189425, "title": "A re-look at MoA data!", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 180500, "title": "This is not Multi-class Classification!", "comment_count": 18, "views": 0, "votes": 135 }, { "id": 188329, "title": "Smaller batch sizes helping to converge faster and better score?", "comment_count": 5, "views": 0, "votes": 9 }, { "id": 180918, "title": "Is there a reason why NN are doing better in this comp ?", "comment_count": 36, "views": 0, "votes": 27 }, { "id": 188898, "title": "Submission Problem", "comment_count": 3, "views": 0, "votes": 0 }, { "id": 188783, "title": "Using pre-made features?", "comment_count": 4, "views": 0, "votes": null }, { "id": 188947, "title": "Hello,how to use TPU with tensorflow1 and Keras?", "comment_count": 6, "views": 0, "votes": -1 }, { "id": 188739, "title": "Public Notebooks blending submission.csv", "comment_count": 1, "views": 0, "votes": null }, { "id": 188865, "title": "I can't submit my file!", "comment_count": 3, "views": 0, "votes": 0 }, { "id": 188667, "title": "Private Leaderboard", "comment_count": 10, "views": 0, "votes": -4 }, { "id": 188704, "title": "is this overfit?", "comment_count": 3, "views": 0, "votes": -1 }, { "id": 188416, "title": "Isn't this imbalanced data？", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 188702, "title": "ML Reproducibility Challenge 2020", "comment_count": 1, "views": 0, "votes": null }, { "id": 180429, "title": "Idea about how to use train_targets_nonscored", "comment_count": 12, "views": 0, "votes": 43 }, { "id": 186717, "title": "I have a question about L1 or L2 regularization in keras.", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 187949, "title": "Optimizing For Poisson Loss works better", "comment_count": 13, "views": 0, "votes": 5 }, { "id": 188273, "title": "How to use dependencies in submission notebook?", "comment_count": 7, "views": 0, "votes": null }, { "id": 187529, "title": "Blogs For Learning Ensembling !", "comment_count": 5, "views": 0, "votes": 8 }, { "id": 187815, "title": "What should be the conditions for blending models?", "comment_count": 7, "views": 0, "votes": 6 }, { "id": 186882, "title": "List of Important Research Papers for Competition", "comment_count": 12, "views": 0, "votes": 19 }, { "id": 188501, "title": "download trained TF model", "comment_count": 1, "views": 0, "votes": null }, { "id": 188515, "title": "One vs. All for this Multilabel Problem", "comment_count": 1, "views": 0, "votes": null }, { "id": 188235, "title": "Notebook Doubt", "comment_count": 5, "views": 0, "votes": null }, { "id": 180588, "title": "Clarifications about the nature of the trainining set data", "comment_count": 8, "views": 0, "votes": 11 }, { "id": 187101, "title": "MoA. genes", "comment_count": 2, "views": 0, "votes": 3 }, { "id": 187586, "title": "is PCA useful?", "comment_count": 6, "views": 0, "votes": 5 }, { "id": 187724, "title": "Log loss - Dumb model - LB Score 0.023", "comment_count": 3, "views": 0, "votes": 6 }, { "id": 187898, "title": "Understanding Cells and Genes", "comment_count": 4, "views": 0, "votes": 3 }, { "id": 187684, "title": "Beginner's question about final submission", "comment_count": 6, "views": 0, "votes": 3 }, { "id": 188363, "title": "What is your best method?", "comment_count": 3, "views": 0, "votes": null }, { "id": 188122, "title": "Beginner notebook suggestion", "comment_count": 3, "views": 0, "votes": 0 }, { "id": 186714, "title": "Deep Gradient Boosted Learning?", "comment_count": 7, "views": 0, "votes": 20 }, { "id": 187983, "title": "Traget encoding", "comment_count": 1, "views": 0, "votes": 2 }, { "id": 187871, "title": "Import from Google Drive: a serious issue", "comment_count": 12, "views": 0, "votes": 2 }, { "id": 188253, "title": "How to get notification of newly posted kernels ?", "comment_count": 2, "views": 0, "votes": null }, { "id": 185989, "title": "Let's guess how much the final best score range will be.", "comment_count": 12, "views": 0, "votes": 8 }, { "id": 187924, "title": "GPU for submissions using Quick Version", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 188159, "title": "MoA Normalization Techniques (Math or Science Geeks Welcome)", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 187108, "title": "2 Hrs to compute public score", "comment_count": 10, "views": 0, "votes": 2 }, { "id": 181113, "title": "Papers on Drug Discovery and Gene Expression", "comment_count": 3, "views": 0, "votes": 26 }, { "id": 181007, "title": "Time to rethink this competition", "comment_count": 31, "views": 0, "votes": 74 }, { "id": 187705, "title": "How to use \"cp_dose\"and\"cp_time\"", "comment_count": 6, "views": 0, "votes": null }, { "id": 187295, "title": "GBM for each target", "comment_count": 5, "views": 0, "votes": 1 }, { "id": 185873, "title": "This is not a Multilabel Classification problem!", "comment_count": 23, "views": 0, "votes": 25 }, { "id": 186822, "title": "A possible clue for denoising?", "comment_count": 0, "views": 0, "votes": 15 }, { "id": 186691, "title": "Tuning Learning Rate While Doing K fold Cross Validation and Fitting on Entire Dataset", "comment_count": 7, "views": 0, "votes": 3 }, { "id": 186919, "title": "g and c columns form 3 clusters on a 2-D space", "comment_count": 6, "views": 0, "votes": 3 }, { "id": 186765, "title": "Top Features ", "comment_count": 3, "views": 0, "votes": 6 }, { "id": 186937, "title": "How to use samples with cp_type \"ctl_vehicle\"?", "comment_count": 6, "views": 0, "votes": 6 }, { "id": 183322, "title": "Per epoch training samples much lower on Kaggle compared to local machine", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 187321, "title": "Failed to submit CSV", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 187226, "title": "How to evaluate models before training it for multiple folds?", "comment_count": 3, "views": 0, "votes": null }, { "id": 186889, "title": "MultilabelStratifiedKFold", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 187477, "title": "pos_weight for nn.BCEWithLogitsLoss", "comment_count": 2, "views": 0, "votes": 0 }, { "id": 180366, "title": "tabular competitions are back? :)", "comment_count": 6, "views": 0, "votes": 9 }, { "id": 187432, "title": "Trouble with log_loss() and StratifiedKFold()", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 181766, "title": "Regularized Greedy Forest (RGF) classifier", "comment_count": 6, "views": 0, "votes": 10 }, { "id": 186643, "title": "What are the advantages of multiclass over multilabel?", "comment_count": 1, "views": 0, "votes": 5 }, { "id": 180733, "title": "Training on nonscore data and finetune on score data helps?", "comment_count": 1, "views": 0, "votes": 7 }, { "id": 186100, "title": "no output file found", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 186426, "title": "Complete newbie to biopharmaceutics, so I don't quite understand the problem we are addressing.", "comment_count": 2, "views": 0, "votes": 3 }, { "id": 181581, "title": "Experiment with Neural Oblivious Decision Ensembles", "comment_count": 2, "views": 0, "votes": 11 }, { "id": 186719, "title": "Discrepancy between CV and LB", "comment_count": 1, "views": 0, "votes": 2 }, { "id": 186260, "title": "Naive approach", "comment_count": 5, "views": 0, "votes": 3 }, { "id": 183835, "title": "Failed Submissions Count", "comment_count": 17, "views": 0, "votes": 2 }, { "id": 186370, "title": "To predict what ? ", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 186350, "title": "Submission error", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 180095, "title": "Wow a competition where the metric is logloss!", "comment_count": 18, "views": 0, "votes": 60 }, { "id": 184018, "title": "Keras Starter notebook for beginners(easy to understand)", "comment_count": 4, "views": 0, "votes": 11 }, { "id": 182274, "title": "A simple explanation of Log Loss metric", "comment_count": 10, "views": 0, "votes": 24 }, { "id": 186018, "title": "Question on the time limits ...", "comment_count": 6, "views": 0, "votes": 2 }, { "id": 185396, "title": "Anyone using Random Forests? ", "comment_count": 9, "views": 0, "votes": null }, { "id": 185728, "title": "Beware of leakage in your local CVs.", "comment_count": 3, "views": 0, "votes": 7 }, { "id": 180330, "title": "Live pair programming on this competition with Abhishek Thakur tomorrow!", "comment_count": 40, "views": 0, "votes": 92 }, { "id": 180061, "title": "It seems we will be competing for each 0.001 soon. Maybe show 5 points after zero?", "comment_count": 16, "views": 0, "votes": 50 }, { "id": 180165, "title": "Don't put drugs into Vehicle Control", "comment_count": 3, "views": 0, "votes": 89 }, { "id": 185559, "title": "Clustering & t-SNE of gene features", "comment_count": 4, "views": 0, "votes": 6 }, { "id": 185701, "title": "Ideas on solving for cyclooxygenase_inhibitor and the like!", "comment_count": 1, "views": 0, "votes": 4 }, { "id": 183355, "title": "Useful Notebooks and info from OSIC competition", "comment_count": 4, "views": 0, "votes": 6 }, { "id": 186341, "title": "Output file not showing", "comment_count": 3, "views": 0, "votes": 0 }, { "id": 180795, "title": "Does Setting ctl_vehicle = 0 improves score?", "comment_count": 5, "views": 0, "votes": 6 }, { "id": 180408, "title": "Ways to reduce notebook time (Is the dataset changed during test?)", "comment_count": 5, "views": 0, "votes": 2 }, { "id": 185848, "title": "CV loss confusion", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 185892, "title": "how to improve results", "comment_count": 5, "views": 0, "votes": 1 }, { "id": 183327, "title": "Support Vector Machines FTW", "comment_count": 18, "views": 0, "votes": 8 }, { "id": 181551, "title": "Any luck with PCA components?", "comment_count": 11, "views": 0, "votes": 8 }, { "id": 181340, "title": "CV vs LB (Single Model - fold 1 or n) ", "comment_count": 12, "views": 0, "votes": 10 }, { "id": 185100, "title": "What is the difference between each SVM option?", "comment_count": 4, "views": 0, "votes": 7 }, { "id": 186011, "title": "Predictions in regards to the control group", "comment_count": 2, "views": 0, "votes": null }, { "id": 186043, "title": "Cellular Response is Complicated", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 185839, "title": "Submission file", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 180737, "title": "g- columns are making some Interesting Curve.", "comment_count": 11, "views": 0, "votes": 23 }, { "id": 183589, "title": "Catboost with MultiRegression Loss", "comment_count": 8, "views": 0, "votes": 13 }, { "id": 186014, "title": "Individual Models vs Positive Counts", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 185567, "title": "Loss, CV Score and Leaderboard score ", "comment_count": 3, "views": 0, "votes": null }, { "id": 185126, "title": "And now for something completely different", "comment_count": 2, "views": 0, "votes": 4 }, { "id": 183980, "title": "What Kind of Drugs are for what kind of research ?", "comment_count": 2, "views": 0, "votes": 5 }, { "id": 185183, "title": "A suggestion to Kaggle Team on daily submissions", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 183670, "title": "Getting Submission Scoring Error", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 185634, "title": "On the multi-label regression problem in Skorch", "comment_count": 3, "views": 0, "votes": null }, { "id": 185083, "title": "A doubt about code competition", "comment_count": 7, "views": 0, "votes": 2 }, { "id": 183553, "title": "How to choose NN hyperparameters?", "comment_count": 7, "views": 0, "votes": 7 }, { "id": 183577, "title": "State of the arts : Architectures to be explored", "comment_count": 0, "views": 0, "votes": 10 }, { "id": 185317, "title": "Is the loss being calculated right? whichever technique i try the loss is always 0.03 higher", "comment_count": 4, "views": 0, "votes": 0 }, { "id": 183435, "title": "Score Bug?", "comment_count": 8, "views": 0, "votes": 4 }, { "id": 184012, "title": "Hello ,there is something wrong.", "comment_count": 5, "views": 0, "votes": null }, { "id": 183485, "title": "Validation issues", "comment_count": 2, "views": 0, "votes": 5 }, { "id": 183806, "title": "Facing issue while submission", "comment_count": 8, "views": 0, "votes": null }, { "id": 182333, "title": "disabling internet during notebook submission", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 180946, "title": "correlation of targets", "comment_count": 2, "views": 0, "votes": 5 }, { "id": 183388, "title": "any ideas how approach the problem are welcome", "comment_count": 9, "views": 0, "votes": null }, { "id": 182626, "title": "Multi-label Classification in Python", "comment_count": 7, "views": 0, "votes": 8 }, { "id": 182792, "title": "Model ensembling", "comment_count": 9, "views": 0, "votes": 10 }, { "id": 184034, "title": "Numbering of notebooks is wrong during submission.", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 183615, "title": "Selective Overfitting", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 183729, "title": "Domain Knowledge", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 183166, "title": "PCA - Additional Features", "comment_count": 4, "views": 0, "votes": 4 }, { "id": 181835, "title": "TabNet and how to go about it", "comment_count": 14, "views": 0, "votes": 17 }, { "id": 183627, "title": "Loss not Improving", "comment_count": 5, "views": 0, "votes": 0 }, { "id": 183419, "title": "why Binary cross-entropy ", "comment_count": 6, "views": 0, "votes": 1 }, { "id": 183896, "title": "how to import datasets in Kaggles notebook environment ", "comment_count": 2, "views": 0, "votes": null }, { "id": 181306, "title": "Multioutput modeling", "comment_count": 15, "views": 0, "votes": 14 }, { "id": 181501, "title": "No internet? How to install R/Bioconductor packages?", "comment_count": 4, "views": 0, "votes": null }, { "id": 181849, "title": "Let's do some EDA", "comment_count": 4, "views": 0, "votes": 4 }, { "id": 183285, "title": "Submission unabled", "comment_count": 5, "views": 0, "votes": 0 }, { "id": 183214, "title": "Code Competition", "comment_count": 4, "views": 0, "votes": null }, { "id": 183326, "title": "GPU Time is increasing in high (double) rate?", "comment_count": 2, "views": 0, "votes": null }, { "id": 182854, "title": "Random Forest (RF) Taking Too Long? ", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 182936, "title": "My submission give Error - Please help", "comment_count": 6, "views": 0, "votes": 2 }, { "id": 181228, "title": "Some OOF Improvement With Gaussian Noise Augmentation", "comment_count": 5, "views": 0, "votes": 34 }, { "id": 183387, "title": "GPU time limit", "comment_count": 4, "views": 0, "votes": null }, { "id": 181418, "title": "How to optimize your PyTorch code", "comment_count": 6, "views": 0, "votes": 38 }, { "id": 180114, "title": "Cannot submit predictions due to bug. \"3 out of 3 submissions remaining\", \"3 out of 3 submissions used\" ", "comment_count": 3, "views": 0, "votes": 9 }, { "id": 182104, "title": "TabNetMultiTaskClassfier Tutorial", "comment_count": 7, "views": 0, "votes": 5 }, { "id": 183334, "title": "how to convert model.predict_proba into the submission form?", "comment_count": 2, "views": 0, "votes": null }, { "id": 183109, "title": "A couple of Questions", "comment_count": 2, "views": 0, "votes": 0 }, { "id": 183352, "title": "Submission completion notifications", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 182443, "title": "Data or Dogma - Downvoting a sincere question?", "comment_count": 4, "views": 0, "votes": 5 }, { "id": 183134, "title": "train_targets_nonscored", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 182284, "title": "High correlation b/w c- features ??", "comment_count": 3, "views": 0, "votes": 4 }, { "id": 181786, "title": "Is there a baseline from organizers?", "comment_count": 4, "views": 0, "votes": 14 }, { "id": 182553, "title": "when submitting notebook running forever, till yesterday it was fine", "comment_count": 1, "views": 0, "votes": 0 }, { "id": 182475, "title": "Bayesian Optimization evaluator function suggestions", "comment_count": 0, "views": 0, "votes": 4 }, { "id": 182396, "title": "List of Useful Blogs and Competitions to Learn About Multi-Label Classification", "comment_count": 0, "views": 0, "votes": 5 }, { "id": 181010, "title": "Questions to the Hosts about Control Perturbations ", "comment_count": 6, "views": 0, "votes": 16 }, { "id": 182791, "title": "A strange question", "comment_count": 2, "views": 0, "votes": 0 }, { "id": 182485, "title": "PCA and Neural Networks", "comment_count": 0, "views": 0, "votes": 3 }, { "id": 182476, "title": "0.018* scores", "comment_count": 4, "views": 0, "votes": null }, { "id": 182539, "title": "Cell lines and genes info , and other data ", "comment_count": 1, "views": 0, "votes": 1 }, { "id": 182806, "title": "Submission error without message", "comment_count": 2, "views": 0, "votes": null }, { "id": 182468, "title": "Cell viability values", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 182345, "title": "Submission score error", "comment_count": 8, "views": 0, "votes": 1 }, { "id": 181326, "title": "[domain] gene expression vs cell type", "comment_count": 7, "views": 0, "votes": 8 }, { "id": 181979, "title": "what's the point of hiding gene names?", "comment_count": 3, "views": 0, "votes": 3 }, { "id": 180981, "title": "3 models by cp_time (24, 48 and 72) instead of one.", "comment_count": 9, "views": 0, "votes": 19 }, { "id": 181798, "title": "c-0~99 Correlation are high", "comment_count": 1, "views": 0, "votes": 9 }, { "id": 182195, "title": "target values more than training samples", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 181968, "title": "Why is accuracy not used in most notebooks?", "comment_count": 9, "views": 0, "votes": 3 }, { "id": 181244, "title": "LightGBM hyperparameter tuning", "comment_count": 6, "views": 0, "votes": 6 }, { "id": 180382, "title": "For enthusiastic beginners: A gentle introduction for those who are new to the domain", "comment_count": 13, "views": 0, "votes": 9 }, { "id": 182373, "title": "any tips to improve the DNN and any good site for hyperopt implementation in NN", "comment_count": 2, "views": 0, "votes": null }, { "id": 182236, "title": "Keras Multilabel Neural Network v3.0 | Starters", "comment_count": 0, "views": 0, "votes": 3 }, { "id": 180309, "title": "What is the biological meaning of rows in dataset ? (if drugs, why 23k >> 5k) ", "comment_count": 6, "views": 0, "votes": 11 }, { "id": 181976, "title": "train_targets_nonscored", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 181813, "title": "Some Labels are unique", "comment_count": 2, "views": 0, "votes": 4 }, { "id": 181308, "title": "External Data Thread ?", "comment_count": 5, "views": 0, "votes": 11 }, { "id": 180695, "title": "Agonist vs Antagonist", "comment_count": 5, "views": 0, "votes": 22 }, { "id": 180311, "title": "40% unknown and 99.66% highly-imbalanced data (in training and test)", "comment_count": 3, "views": 0, "votes": 26 }, { "id": 180554, "title": "LGBM approach (206 models for each target)", "comment_count": 13, "views": 0, "votes": 37 }, { "id": 182259, "title": "How to handle classes with just one positive id? ", "comment_count": 1, "views": 0, "votes": 2 }, { "id": 181759, "title": "ModuleNotFoundError: No module named 'pyod'?", "comment_count": 2, "views": 0, "votes": null }, { "id": 181263, "title": "\"Verbosity\" - public service announcement", "comment_count": 6, "views": 0, "votes": 4 }, { "id": 181389, "title": "A few helpful concept videos", "comment_count": 1, "views": 0, "votes": 10 }, { "id": 181634, "title": "Different subsets of features for different targets", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 182401, "title": "Not Able to Submit a Particular Submission File", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 182400, "title": "Huge difference in CV and LB scores?", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 181549, "title": "Wanted: domain knowledge source", "comment_count": 1, "views": 0, "votes": 6 }, { "id": 182379, "title": "target encoding", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 181322, "title": "📊 Statistical Analysis 📈", "comment_count": 2, "views": 0, "votes": 9 }, { "id": 181709, "title": "Anyone using train_targets_nonscored dataset to improve the predictions?", "comment_count": 1, "views": 0, "votes": 5 }, { "id": 182152, "title": "Submission error when adding files from private dataset?", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 181840, "title": "random seed", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 182344, "title": " X data contains of gene expression and cell viability columns. we need to predict the which drug supports for MOA !!!.", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 182318, "title": "Model per label approach question", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 181357, "title": "How many \"baselines\" are there in this competition?", "comment_count": 2, "views": 0, "votes": 7 }, { "id": 181744, "title": "A review of connectivity map", "comment_count": 0, "views": 0, "votes": 3 }, { "id": 181352, "title": "Any hints to beat the 0.01862 ?🔥", "comment_count": 4, "views": 0, "votes": 3 }, { "id": 181716, "title": "Make sure you check your validation classes!", "comment_count": 3, "views": 0, "votes": 2 }, { "id": 181663, "title": "No Output file found", "comment_count": 3, "views": 0, "votes": 3 }, { "id": 181339, "title": "Out-of-Fold (a.k.a. OOF ) Score", "comment_count": 0, "views": 0, "votes": 7 }, { "id": 181951, "title": "Pip install AutoGLUon successful but Import Error", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 180389, "title": "cp_type column data", "comment_count": 3, "views": 0, "votes": 7 }, { "id": 181738, "title": "Column Meanings ", "comment_count": 5, "views": 0, "votes": null }, { "id": 181743, "title": "gene data with -10 values", "comment_count": 3, "views": 0, "votes": null }, { "id": 181745, "title": "Help for beginners.", "comment_count": 0, "views": 0, "votes": 2 }, { "id": 181946, "title": "New Notebook (Kernel) for this competition", "comment_count": 3, "views": 0, "votes": -2 }, { "id": 180140, "title": "Experience Submission Scoring Error", "comment_count": 9, "views": 0, "votes": 3 }, { "id": 180656, "title": "Missing labels in the dataset", "comment_count": 7, "views": 0, "votes": 13 }, { "id": 181822, "title": "All Questions about the Mechanisms of Action (MoA) Prediction", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 180142, "title": "Machine Learning vs Deep Learning", "comment_count": 12, "views": 0, "votes": 14 }, { "id": 181896, "title": "What does cp_time actually represent ?", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 180660, "title": "Stratified KFold on training dataset", "comment_count": 2, "views": 0, "votes": 2 }, { "id": 180098, "title": "The meaning of gene expression and cell viability features?", "comment_count": 5, "views": 0, "votes": 27 }, { "id": 181104, "title": "Notebook Timeout", "comment_count": 3, "views": 0, "votes": 3 }, { "id": 180756, "title": "What g-0, g-1,..... signify?", "comment_count": 1, "views": 0, "votes": 5 }, { "id": 180714, "title": "Which approach is better? ", "comment_count": 4, "views": 0, "votes": 3 }, { "id": 181250, "title": "Simple explanations for Gene Expression, Cell Viability & Mechanism of Actions?", "comment_count": 3, "views": 0, "votes": null }, { "id": 180123, "title": "Possible Glitch with UTC reset...", "comment_count": 2, "views": 0, "votes": 5 }, { "id": 181052, "title": "Availability of drug names", "comment_count": 1, "views": 0, "votes": 3 }, { "id": 181294, "title": "I have a question about My accuracy.", "comment_count": 4, "views": 0, "votes": 0 }, { "id": 180010, "title": "Connectivity Map", "comment_count": 2, "views": 0, "votes": 28 }, { "id": 181076, "title": "Metrics in the Drug Discovery Context", "comment_count": 0, "views": 0, "votes": 3 }, { "id": 181552, "title": "Rethink loss function. Is top-K error better suited?", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 181553, "title": "Rethink loss function. Is top-K error better suited?", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 181120, "title": "Dose - D1=low D2=high?", "comment_count": 1, "views": 0, "votes": 2 }, { "id": 180927, "title": "Help! Submission Scoring Error Reasons Thread", "comment_count": 1, "views": 0, "votes": 3 }, { "id": 180417, "title": "Examining the gap between LB and validation", "comment_count": 6, "views": 0, "votes": 8 }, { "id": 181227, "title": "Multi Log loss function for Neural Network - MoA", "comment_count": 0, "views": 0, "votes": 2 }, { "id": 181329, "title": "CatBoost, LightGBM and Ensemble models", "comment_count": 7, "views": 0, "votes": 1 }, { "id": 181472, "title": "How to create the submission file?", "comment_count": 1, "views": 0, "votes": null }, { "id": 180841, "title": "How can I feed my model with the entire training data", "comment_count": 5, "views": 0, "votes": null }, { "id": 181183, "title": "Using class names as feature", "comment_count": 0, "views": 0, "votes": 2 }, { "id": 180744, "title": "Some materials on kaggle related to genes expressions", "comment_count": 0, "views": 0, "votes": 5 }, { "id": 180058, "title": "First Public Notebook, please check it out!", "comment_count": 14, "views": 0, "votes": 8 }, { "id": 180427, "title": "What is the nonscored csv file?", "comment_count": 5, "views": 0, "votes": null }, { "id": 180159, "title": "Public and Private Test Sets", "comment_count": 2, "views": 0, "votes": 13 }, { "id": 180329, "title": "Genes expressions corresponds to single-cell/averaged over cell type/averaged over all cells ? ", "comment_count": 0, "views": 0, "votes": 11 }, { "id": 181279, "title": "Inference Time : NN vs Traditional Appraoches", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 180436, "title": "Submissions are not resetting again.", "comment_count": 1, "views": 0, "votes": 5 }, { "id": 180375, "title": "Experiment Design ", "comment_count": 0, "views": 0, "votes": 8 }, { "id": 181127, "title": "Confusing Labels", "comment_count": 5, "views": 0, "votes": 1 }, { "id": 180251, "title": "Fast submission", "comment_count": 0, "views": 0, "votes": 9 }, { "id": 180187, "title": "Intuition of Evaluation Metrics", "comment_count": 5, "views": 0, "votes": 4 }, { "id": 180451, "title": "Setting up a good validation scheme", "comment_count": 4, "views": 0, "votes": 2 }, { "id": 181053, "title": "Availability of drug names", "comment_count": 0, "views": 0, "votes": 0 }, { "id": 180891, "title": "Data Prep for Target class", "comment_count": 3, "views": 0, "votes": null }, { "id": 180134, "title": "Continuous numerical features with Gaussian-like distributions and outliers", "comment_count": 2, "views": 0, "votes": 6 }, { "id": 180640, "title": "target variables", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 180072, "title": "Mapping of the genes to their names", "comment_count": 1, "views": 0, "votes": 6 }, { "id": 180353, "title": "New to Multi-LABEL classification task. What to know and what not to miss?", "comment_count": 4, "views": 0, "votes": 1 }, { "id": 180367, "title": "No internet access - no external python modules", "comment_count": 3, "views": 0, "votes": 1 }, { "id": 180648, "title": "Can sig_id agents in the training set have undiscovered properties?", "comment_count": 1, "views": 0, "votes": null }, { "id": 180590, "title": "Features c- and g- meaning", "comment_count": 0, "views": 0, "votes": 1 }, { "id": 180232, "title": "can't complete course", "comment_count": 2, "views": 0, "votes": 1 }, { "id": 180441, "title": "Interesting competetion : Coronavirus Drug Discovery", "comment_count": 0, "views": 0, "votes": 0 } ], "errors": [] }