{
  "total": 148,
  "topics": [
    {
      "id": 304454,
      "title": "New to Machine Learning or to Kaggle? Check this out.",
      "comment_count": 2,
      "views": 0,
      "votes": 19
    },
    {
      "id": 304455,
      "title": "Looking for a Team Megathread",
      "comment_count": 17,
      "views": 0,
      "votes": 16
    },
    {
      "id": 304814,
      "title": "I would have never imagined that ERT would have gotten so successful in this competition",
      "comment_count": 11,
      "views": 0,
      "votes": 28
    },
    {
      "id": 306172,
      "title": "Neural Nets for Tabular Data",
      "comment_count": 2,
      "views": 0,
      "votes": 17
    },
    {
      "id": 304561,
      "title": "Clusters … UMAP, T-SNE, LDA",
      "comment_count": 5,
      "views": 0,
      "votes": 13
    },
    {
      "id": 310359,
      "title": "#1 Solution: Exploiting the Flawed Random Generation",
      "comment_count": 16,
      "views": 0,
      "votes": 70
    },
    {
      "id": 306855,
      "title": "BioPython - nice way to operate on DNA sequence",
      "comment_count": 9,
      "views": 0,
      "votes": 33
    },
    {
      "id": 305431,
      "title": "How to avoid burnout?",
      "comment_count": 34,
      "views": 0,
      "votes": 36
    },
    {
      "id": 306614,
      "title": "Google Colab | Instead of Kaggle notebook",
      "comment_count": 32,
      "views": 0,
      "votes": 44
    },
    {
      "id": 305079,
      "title": "a cool interactive website to understand probability and statistics.",
      "comment_count": 13,
      "views": 0,
      "votes": 29
    },
    {
      "id": 305350,
      "title": "Introducing Nested Cross-Validation = \"Double Cross-Validation\"",
      "comment_count": 9,
      "views": 0,
      "votes": 55
    },
    {
      "id": 308045,
      "title": "Permutation importance runtime",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 309488,
      "title": "Congratulations to the new Remek Kinas - new Discussion Grandmaster! 🎉",
      "comment_count": 35,
      "views": 0,
      "votes": 21
    },
    {
      "id": 310367,
      "title": "#2 solution",
      "comment_count": 7,
      "views": 0,
      "votes": 37
    },
    {
      "id": 305364,
      "title": "Save the earth one tiny bit by dropping 38% of training data with duplicated rows",
      "comment_count": 33,
      "views": 0,
      "votes": 74
    },
    {
      "id": 310244,
      "title": "That last day of competition feeling",
      "comment_count": 6,
      "views": 0,
      "votes": 10
    },
    {
      "id": 310373,
      "title": "Takeaways from the TPS February 2022",
      "comment_count": 3,
      "views": 0,
      "votes": 13
    },
    {
      "id": 310461,
      "title": "Denoising Autoencoder- Unique solution for TPS-FEB2022",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 309939,
      "title": "Data Visualization cheat sheet by Dr. Andrew Abela",
      "comment_count": 0,
      "views": 0,
      "votes": 30
    },
    {
      "id": 310507,
      "title": "Takeaways Tabular Playground series Feb 22",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 310108,
      "title": "Looking Forward to Solution Presentations",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 308950,
      "title": "Heartiest Congratulations to @vad13irt  on becoming the Discussion Grandmaster",
      "comment_count": 23,
      "views": 0,
      "votes": 14
    },
    {
      "id": 309742,
      "title": "Small competition break ",
      "comment_count": 7,
      "views": 0,
      "votes": 10
    },
    {
      "id": 310220,
      "title": "Why are there negative values for any row?",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 309748,
      "title": "Pyramid tips",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 309858,
      "title": "Best Way to Ensemble?",
      "comment_count": 5,
      "views": 0,
      "votes": 3
    },
    {
      "id": 310229,
      "title": "Public Score not linked with Notebook Help",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 308606,
      "title": "The workflow to become a Kaggle Discussion Master :)",
      "comment_count": 7,
      "views": 0,
      "votes": 13
    },
    {
      "id": 310145,
      "title": "Tabular Playground Series - Feb 2022 99.43% Accuracy ",
      "comment_count": 4,
      "views": 0,
      "votes": -6
    },
    {
      "id": 308094,
      "title": "Kaggle tip for beginners",
      "comment_count": 7,
      "views": 0,
      "votes": 17
    },
    {
      "id": 308386,
      "title": "Using a AutoEncoder/Variational AutoEncoder",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 308876,
      "title": "GC_content (improved my score)",
      "comment_count": 6,
      "views": 0,
      "votes": 14
    },
    {
      "id": 308644,
      "title": "Min, Max,Mean, Std: Basic Feature Engineering Question?",
      "comment_count": 7,
      "views": 0,
      "votes": 10
    },
    {
      "id": 308827,
      "title": "Should this be a rare case to trust the public leaderboard? ",
      "comment_count": 5,
      "views": 0,
      "votes": 6
    },
    {
      "id": 307328,
      "title": "Beginner friendly compilation of good starter notebooks",
      "comment_count": 8,
      "views": 0,
      "votes": 28
    },
    {
      "id": 308637,
      "title": "Which batch size should I use?",
      "comment_count": 6,
      "views": 0,
      "votes": 2
    },
    {
      "id": 308754,
      "title": "Effect of N number of PCA components on Accuracy ",
      "comment_count": 9,
      "views": 0,
      "votes": 10
    },
    {
      "id": 308645,
      "title": "What I've learned from this compitition.",
      "comment_count": 10,
      "views": 0,
      "votes": 19
    },
    {
      "id": 307038,
      "title": "State of the Art Neural Networks for Tabular Deep Learning",
      "comment_count": 14,
      "views": 0,
      "votes": 30
    },
    {
      "id": 308312,
      "title": "🎯100% Accuracy🎯 on at least 50% of the Data guarantee (NO MACHINE LEARNING)",
      "comment_count": 14,
      "views": 0,
      "votes": 8
    },
    {
      "id": 305184,
      "title": "Kaggle is  Addictive, how do you dophamine detox and became effective on kaggle?",
      "comment_count": 53,
      "views": 0,
      "votes": 73
    },
    {
      "id": 305040,
      "title": "New Kaggle Notebooks Grandmaster, Sanskar Hasija! ",
      "comment_count": 42,
      "views": 0,
      "votes": 38
    },
    {
      "id": 308462,
      "title": "How to perform Error Analysis",
      "comment_count": 6,
      "views": 0,
      "votes": 4
    },
    {
      "id": 308207,
      "title": "Why the duplicate dataset performs better than the non duplicate dataset?",
      "comment_count": 6,
      "views": 0,
      "votes": 8
    },
    {
      "id": 307272,
      "title": "Speed up Random Forest",
      "comment_count": 8,
      "views": 0,
      "votes": 15
    },
    {
      "id": 308425,
      "title": "Introducing and Implementing  PyTorch_Tabular",
      "comment_count": 7,
      "views": 0,
      "votes": 4
    },
    {
      "id": 308183,
      "title": "Visualization of Escherichia_coli with different error rate",
      "comment_count": 1,
      "views": 0,
      "votes": 7
    },
    {
      "id": 308465,
      "title": "Handling large datasets",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 307367,
      "title": "Multiple models / Blending / Bagging / Boosting Ensemble - my TPS-02 tutorials",
      "comment_count": 13,
      "views": 0,
      "votes": 25
    },
    {
      "id": 308297,
      "title": "Any suggestion to improvement the score?",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 307922,
      "title": "Why is Extra Trees Classifier the best chcoice for this competition?",
      "comment_count": 8,
      "views": 0,
      "votes": 4
    },
    {
      "id": 307132,
      "title": "TPS - Feb 2022, What I have tried.",
      "comment_count": 19,
      "views": 0,
      "votes": 17
    },
    {
      "id": 308618,
      "title": "A lot features",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 304745,
      "title": "I am Kaggle Discussion Master! ",
      "comment_count": 58,
      "views": 0,
      "votes": 69
    },
    {
      "id": 305845,
      "title": "Experiment - bacteria as image classification - Keras Conv2D",
      "comment_count": 30,
      "views": 0,
      "votes": 27
    },
    {
      "id": 308014,
      "title": "Stuck at a certain accuracy",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 308180,
      "title": "Tried PCA, very nice for EDA insight & processing, but lost 1.5pp accuracy",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 307266,
      "title": "Implementing Nested Cross-Validation for TPS Feb 2022",
      "comment_count": 10,
      "views": 0,
      "votes": 9
    },
    {
      "id": 308274,
      "title": "Deep Neural Networks for Tabular| Data in-depth look| survey paper | 19 pages",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 308210,
      "title": "POST PROCESSING",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 307835,
      "title": "Create classifiers based on GCD values - why I got low score?",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 307681,
      "title": "IPython Features, which you do not know",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 305958,
      "title": "What I have learnt from Paper",
      "comment_count": 13,
      "views": 0,
      "votes": 17
    },
    {
      "id": 307997,
      "title": "How to improve score?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 308079,
      "title": "Implementing TabTransformer └[∵┌] └[ ∵ ]┘[┐∵]┘",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 306775,
      "title": "What should you do?",
      "comment_count": 7,
      "views": 0,
      "votes": 8
    },
    {
      "id": 308005,
      "title": "How to tackle the entries corresponding to GCD 10000 ? ",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 307217,
      "title": "Any tips to create a small statistically significant sample of the original dataset?",
      "comment_count": 4,
      "views": 0,
      "votes": 5
    },
    {
      "id": 307497,
      "title": "In many notebooks people keep adjusting probability to keep ~10% for each bacteria in predicted data",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 307608,
      "title": "Feature Engineering- Only 4 Features - A, T, C, G",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 305870,
      "title": "We do not need CV with ExtraTreeClassifier (ETC)",
      "comment_count": 16,
      "views": 0,
      "votes": 16
    },
    {
      "id": 306581,
      "title": "Papers with Code: Deep Learning for Tabular Data",
      "comment_count": 5,
      "views": 0,
      "votes": 17
    },
    {
      "id": 305713,
      "title": "Yes! you can use GPU for faster prediction for Scikit-learn & XGBoost models (using Hummingbird-ml)",
      "comment_count": 5,
      "views": 0,
      "votes": 8
    },
    {
      "id": 307759,
      "title": "Correctly sample a huge dataset - TPS Feb 22",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 307275,
      "title": " FastAI Model - Turn it into a Toon!!! TPS Feb 22 - FastAI Model",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 307690,
      "title": "raw files from \"my submissions\" history with scores",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 306701,
      "title": "Discussion on KNN: Neighbours=1?",
      "comment_count": 5,
      "views": 0,
      "votes": 5
    },
    {
      "id": 307535,
      "title": "Lost Notebook - HELP",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 305234,
      "title": "KNN ? Why long time taken??",
      "comment_count": 16,
      "views": 0,
      "votes": 3
    },
    {
      "id": 307032,
      "title": "Was anyone able to reproduce sampling figures like the ones in the paper? (Solved, Made a wrapper around the author's plotting function))",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 305343,
      "title": "Create a new IMAGE dataset from the data ",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 306580,
      "title": "About LB scores",
      "comment_count": 9,
      "views": 0,
      "votes": 4
    },
    {
      "id": 306761,
      "title": "How to select the classifier? ",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 306684,
      "title": "Some samples from the test dataset exists in the train dataset !",
      "comment_count": 4,
      "views": 0,
      "votes": 6
    },
    {
      "id": 305276,
      "title": "New Kaggle Notebooks Master, Devlikamov Vlad!",
      "comment_count": 20,
      "views": 0,
      "votes": 17
    },
    {
      "id": 305733,
      "title": "Reconsider whether duplicates should be dropped",
      "comment_count": 20,
      "views": 0,
      "votes": 23
    },
    {
      "id": 306620,
      "title": "Combining folds from different models",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 305933,
      "title": "No need of feature scaling with decision-trees",
      "comment_count": 9,
      "views": 0,
      "votes": 5
    },
    {
      "id": 306435,
      "title": "Deduped Data with Stratified Folds",
      "comment_count": 0,
      "views": 0,
      "votes": 11
    },
    {
      "id": 306072,
      "title": "How duplicate rows came about in the original paper",
      "comment_count": 9,
      "views": 0,
      "votes": 11
    },
    {
      "id": 305751,
      "title": "Managing Common Python Code?",
      "comment_count": 6,
      "views": 0,
      "votes": 8
    },
    {
      "id": 306021,
      "title": "New Design of Leaderboard! ",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 305716,
      "title": "What did you try but didn't work?",
      "comment_count": 13,
      "views": 0,
      "votes": 10
    },
    {
      "id": 305516,
      "title": "Data quality alert",
      "comment_count": 12,
      "views": 0,
      "votes": 19
    },
    {
      "id": 306162,
      "title": "Correlation between bacteria's of same families!!",
      "comment_count": 3,
      "views": 0,
      "votes": 8
    },
    {
      "id": 305950,
      "title": "Taxonomy of the bacterial species pulled from the NCBI database to show how the species are related",
      "comment_count": 6,
      "views": 0,
      "votes": 10
    },
    {
      "id": 304949,
      "title": "What's your Best CV scores?",
      "comment_count": 11,
      "views": 0,
      "votes": 13
    },
    {
      "id": 305060,
      "title": "Data science = observing + modeling",
      "comment_count": 12,
      "views": 0,
      "votes": 28
    },
    {
      "id": 305891,
      "title": "XGBoost Feature Importance Across Folds",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 305319,
      "title": "No opportunity to reproduce the results!",
      "comment_count": 9,
      "views": 0,
      "votes": 14
    },
    {
      "id": 304472,
      "title": "Start with domain knowledge - nature of problem - BACTERIAS in competition",
      "comment_count": 10,
      "views": 0,
      "votes": 23
    },
    {
      "id": 306163,
      "title": "PseudoLabelling increases CV by 0.00653 and LB by 0.00015",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 306299,
      "title": "Submission Questions",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 305750,
      "title": "🚀Dimensionality Reduction Techniques: A to Z",
      "comment_count": 7,
      "views": 0,
      "votes": 5
    },
    {
      "id": 305463,
      "title": "Surprising results regarding n_estimators for ExtraTrees & Accuracy",
      "comment_count": 1,
      "views": 0,
      "votes": 7
    },
    {
      "id": 304813,
      "title": "ExtraTreesClassifier: How Accuracy changes with Number of Estimators",
      "comment_count": 3,
      "views": 0,
      "votes": 16
    },
    {
      "id": 306424,
      "title": "Need more information on the Data (Biologist background appreciated).",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 305409,
      "title": "CV vs LB gap using neural networks",
      "comment_count": 8,
      "views": 0,
      "votes": 13
    },
    {
      "id": 304483,
      "title": "8 Categorical Features ",
      "comment_count": 19,
      "views": 0,
      "votes": 37
    },
    {
      "id": 304459,
      "title": "Speed up your kernel using Intel® Extension for Scikit-learn",
      "comment_count": 8,
      "views": 0,
      "votes": 33
    },
    {
      "id": 305690,
      "title": "Dimensionality Reduction Visualization",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 304457,
      "title": "Hyperparameters Optimization",
      "comment_count": 20,
      "views": 0,
      "votes": 25
    },
    {
      "id": 305634,
      "title": "I see overfitting everywhere",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 305654,
      "title": "OOF predictions vs prediction from fitted model",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 305389,
      "title": "Any good neural network based model?",
      "comment_count": 12,
      "views": 0,
      "votes": 3
    },
    {
      "id": 305020,
      "title": "Reduce train dataset to 11% of initial dataset (234 MB)",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 305645,
      "title": "Why you should be Spot-Checking Algorithms on your Machine Learning Problems",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 304456,
      "title": "Optimization Python Code",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 304988,
      "title": "📌 TPS-Feb22, EDA📊 + Modeling🔥",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 304881,
      "title": "Public LB is just another fold.",
      "comment_count": 6,
      "views": 0,
      "votes": 12
    },
    {
      "id": 305665,
      "title": "TensorFlow Decision Forests",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 305056,
      "title": "This comparison confirms the performance of ExtraTreesClassifier",
      "comment_count": 2,
      "views": 0,
      "votes": 11
    },
    {
      "id": 305053,
      "title": "Extra Trees vs Random Forest explained",
      "comment_count": 1,
      "views": 0,
      "votes": 6
    },
    {
      "id": 304720,
      "title": "Why Extra Trees seems to be outperforming Gradient Booster?",
      "comment_count": 6,
      "views": 0,
      "votes": 11
    },
    {
      "id": 305342,
      "title": "A bit of SWAG encouragement ",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 304616,
      "title": "How the data for this competition was generated",
      "comment_count": 1,
      "views": 0,
      "votes": 31
    },
    {
      "id": 304618,
      "title": "Use the mode when cross-validating",
      "comment_count": 1,
      "views": 0,
      "votes": 10
    },
    {
      "id": 305063,
      "title": "How to Blend Several Models for Multi-Class Classification?",
      "comment_count": 5,
      "views": 0,
      "votes": 5
    },
    {
      "id": 304810,
      "title": "To reproduce exactly all the feature names",
      "comment_count": 4,
      "views": 0,
      "votes": 6
    },
    {
      "id": 305145,
      "title": "Two-way Interaction",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 304617,
      "title": "Bases distribution",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 304458,
      "title": "Tutorials on reading large datasets",
      "comment_count": 2,
      "views": 0,
      "votes": 14
    },
    {
      "id": 304519,
      "title": "Interesting dependency feature names",
      "comment_count": 12,
      "views": 0,
      "votes": 11
    },
    {
      "id": 304507,
      "title": "TARGET Category Distribution ",
      "comment_count": 2,
      "views": 0,
      "votes": 11
    },
    {
      "id": 304571,
      "title": "PCA Analysis",
      "comment_count": 4,
      "views": 0,
      "votes": 9
    },
    {
      "id": 304777,
      "title": "TPS-Feb-2022 - A Brilliant resource for reading large datasets!🏆👍🔮",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 305120,
      "title": "An idea: multi-objective optimisation for feature selection, is this ever used?",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 304513,
      "title": "How to make your model recognize features are closely related?",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 304496,
      "title": "Reduce Memory Usage & Deal with Bulky Data!",
      "comment_count": 2,
      "views": 0,
      "votes": 7
    },
    {
      "id": 304896,
      "title": "What the numbers mean: a simple overview",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 304726,
      "title": "  Big Data Pipelines - How to handle massive data and profile the code",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 304856,
      "title": "Dimensionality Reduction",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 304797,
      "title": "Simple way to load and reduce memory footprint simultaneously",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 304626,
      "title": "What do the numbers mean ?",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 305017,
      "title": "Interupt Optuna training and save??",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 304657,
      "title": "What do negative values mean?",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 305132,
      "title": "ValueError: Shapes (None, 1) and (None, 10) are incompatible",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 304485,
      "title": "Let`s play - Big data mode :) ",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    }
  ],
  "errors": []
}