{
  "total": 139,
  "topics": [
    {
      "id": 291936,
      "title": "Looking for a Team Megathread",
      "comment_count": 12,
      "views": 0,
      "votes": 9
    },
    {
      "id": 291935,
      "title": "New to Machine Learning or to Kaggle? Check this out.",
      "comment_count": 1,
      "views": 0,
      "votes": 14
    },
    {
      "id": 294787,
      "title": "SubTab - Self-Supervised Representation Learning for Tabular Data",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 292839,
      "title": "Pseudo-Labeling ",
      "comment_count": 15,
      "views": 0,
      "votes": 23
    },
    {
      "id": 298304,
      "title": "2# place winning solution.",
      "comment_count": 32,
      "views": 0,
      "votes": 54
    },
    {
      "id": 296842,
      "title": "TPS competitions - my lessons learned in 2021",
      "comment_count": 62,
      "views": 0,
      "votes": 116
    },
    {
      "id": 297960,
      "title": "Overfitting is (sometimes) all you need",
      "comment_count": 13,
      "views": 0,
      "votes": 4
    },
    {
      "id": 291823,
      "title": "Speed up your kernel using Intel® Extension for Scikit-learn",
      "comment_count": 13,
      "views": 0,
      "votes": 58
    },
    {
      "id": 298131,
      "title": "Congratulations, looking forward to reading about your solutions!",
      "comment_count": 12,
      "views": 0,
      "votes": 13
    },
    {
      "id": 298135,
      "title": "Time to say thank you",
      "comment_count": 18,
      "views": 0,
      "votes": 23
    },
    {
      "id": 298050,
      "title": "Happy New Year TPS Kagglers and see you in .... TPS-01-2022 ...",
      "comment_count": 23,
      "views": 0,
      "votes": 24
    },
    {
      "id": 298336,
      "title": "TPS- Dec 2021 - Some Beginner level questions",
      "comment_count": 5,
      "views": 0,
      "votes": 5
    },
    {
      "id": 297834,
      "title": "Quotes, which could change your mind! ",
      "comment_count": 6,
      "views": 0,
      "votes": 10
    },
    {
      "id": 298016,
      "title": "NN: How long does your model take to train?",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 298449,
      "title": "Late Submission",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 297405,
      "title": "Faster Optimization using GPU Multiprocessing",
      "comment_count": 3,
      "views": 0,
      "votes": 11
    },
    {
      "id": 297338,
      "title": "Activation Functions",
      "comment_count": 2,
      "views": 0,
      "votes": 18
    },
    {
      "id": 297584,
      "title": "Optimizing ensemble weights",
      "comment_count": 8,
      "views": 0,
      "votes": 13
    },
    {
      "id": 297614,
      "title": "CodeGrepper: code search engine for accrue code answers as a browser extension.",
      "comment_count": 6,
      "views": 0,
      "votes": 10
    },
    {
      "id": 297903,
      "title": "[Help - LGBM + Optuna]  Kernel execution time too long",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 297919,
      "title": "31st is the last day of this competition!",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 298137,
      "title": "How did blender notebook work in private leaderboard?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 297539,
      "title": "What are your expectations for the closing of this TPS?",
      "comment_count": 6,
      "views": 0,
      "votes": 5
    },
    {
      "id": 297820,
      "title": "Why accuracy is different?",
      "comment_count": 12,
      "views": 0,
      "votes": null
    },
    {
      "id": 293612,
      "title": "Feature engineering update thread",
      "comment_count": 24,
      "views": 0,
      "votes": 47
    },
    {
      "id": 295924,
      "title": ".95696 XGBoost submission",
      "comment_count": 15,
      "views": 0,
      "votes": 22
    },
    {
      "id": 297305,
      "title": "What's the reason ?",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 296567,
      "title": "Implementing SAINT model",
      "comment_count": 15,
      "views": 0,
      "votes": 8
    },
    {
      "id": 297016,
      "title": "LightGBM better than other boosting algos and potentially NN",
      "comment_count": 6,
      "views": 0,
      "votes": 4
    },
    {
      "id": 296912,
      "title": "Read your confusion matrix",
      "comment_count": 8,
      "views": 0,
      "votes": 12
    },
    {
      "id": 296118,
      "title": "Why Does AutoML Matter?",
      "comment_count": 14,
      "views": 0,
      "votes": 28
    },
    {
      "id": 295086,
      "title": "protip:  hoarding knowledge on TPS is poor game theory (Unless...)",
      "comment_count": 5,
      "views": 0,
      "votes": 12
    },
    {
      "id": 296631,
      "title": "Revisiting Deep Learning Models for Tabular Data.",
      "comment_count": 5,
      "views": 0,
      "votes": 13
    },
    {
      "id": 296782,
      "title": "Holiday reading list needed !!!",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 295749,
      "title": "Dropout & BatchNormalization",
      "comment_count": 15,
      "views": 0,
      "votes": 14
    },
    {
      "id": 294089,
      "title": "Boosting methods for multi-class imbalanced data classification.",
      "comment_count": 3,
      "views": 0,
      "votes": 11
    },
    {
      "id": 293878,
      "title": "TPS Dec 2021 :Beginner friendly Baseline compilation",
      "comment_count": 5,
      "views": 0,
      "votes": 24
    },
    {
      "id": 296300,
      "title": "Kaggle code snippets for Google Colab",
      "comment_count": 10,
      "views": 0,
      "votes": 12
    },
    {
      "id": 296516,
      "title": "Tabular Data: Deep Learning is Not All You Need [arXiv+ openreview.net]",
      "comment_count": 5,
      "views": 0,
      "votes": 5
    },
    {
      "id": 292823,
      "title": "Sum of Soil_Type and Wilderness_Area",
      "comment_count": 12,
      "views": 0,
      "votes": 46
    },
    {
      "id": 296075,
      "title": "Hillshade unrelated to Aspect and Slope (But it was in the real data)",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 295934,
      "title": "9 hour timeout",
      "comment_count": 11,
      "views": 0,
      "votes": 6
    },
    {
      "id": 296527,
      "title": "Well-tuned Simple Nets Excel on Tabular Datasets [NeurIPS 2021 paper]",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 295617,
      "title": "Halfway into the competition, the most importance evidence",
      "comment_count": 13,
      "views": 0,
      "votes": 24
    },
    {
      "id": 296039,
      "title": "Dec TPS 92.6% Accuracy in less than 7 mins - Deep NN Model",
      "comment_count": 2,
      "views": 0,
      "votes": 10
    },
    {
      "id": 296751,
      "title": "Validation strategies",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 295957,
      "title": "Motivational quote from François Chollet's book about learning AI",
      "comment_count": 1,
      "views": 0,
      "votes": 12
    },
    {
      "id": 295210,
      "title": "Class distributions for the public test set",
      "comment_count": 7,
      "views": 0,
      "votes": 13
    },
    {
      "id": 296088,
      "title": "Stratified Validation Set instead of Cross-Validation",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 295711,
      "title": "Chunks, label distribution by Id",
      "comment_count": 7,
      "views": 0,
      "votes": 6
    },
    {
      "id": 295849,
      "title": "Getting insight from AutoML",
      "comment_count": 9,
      "views": 0,
      "votes": 5
    },
    {
      "id": 295844,
      "title": "Best single model LB&CV thread",
      "comment_count": 6,
      "views": 0,
      "votes": 5
    },
    {
      "id": 295733,
      "title": "Low Variance Columns",
      "comment_count": 8,
      "views": 0,
      "votes": 3
    },
    {
      "id": 295816,
      "title": "Papers on the uses and limitations of synthetic data",
      "comment_count": 1,
      "views": 0,
      "votes": 11
    },
    {
      "id": 293329,
      "title": "It is time for Pseudolabeling ... ",
      "comment_count": 4,
      "views": 0,
      "votes": 13
    },
    {
      "id": 294847,
      "title": "Planning the ultimate TSP in 2022",
      "comment_count": 21,
      "views": 0,
      "votes": 21
    },
    {
      "id": 295474,
      "title": "How to solve slow running?",
      "comment_count": 6,
      "views": 0,
      "votes": 2
    },
    {
      "id": 295676,
      "title": "Community blending",
      "comment_count": 10,
      "views": 0,
      "votes": 3
    },
    {
      "id": 294775,
      "title": "Error analysis - estimating the effect of pseudo-labelling",
      "comment_count": 8,
      "views": 0,
      "votes": 15
    },
    {
      "id": 292381,
      "title": "Train and test data don't overlap",
      "comment_count": 14,
      "views": 0,
      "votes": 23
    },
    {
      "id": 294305,
      "title": "Imbalanced classes vs. imbalanced cost",
      "comment_count": 21,
      "views": 0,
      "votes": 16
    },
    {
      "id": 295971,
      "title": "GridSearchCV in xgboost: FitFailedWarning: Estimator fit failed",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 294114,
      "title": "Better LB score estimate",
      "comment_count": 5,
      "views": 0,
      "votes": 6
    },
    {
      "id": 295549,
      "title": "Feature Engineering: Encode Soil_Type as Five 8-bit Integers",
      "comment_count": 3,
      "views": 0,
      "votes": 7
    },
    {
      "id": 294845,
      "title": "On using SELU activation",
      "comment_count": 3,
      "views": 0,
      "votes": 25
    },
    {
      "id": 295777,
      "title": "Exemption from quiet period?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 295623,
      "title": "Feature Engineering: Soil Features",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 295026,
      "title": "Feature engineering, use pd.clip function",
      "comment_count": 2,
      "views": 0,
      "votes": 11
    },
    {
      "id": 293373,
      "title": "Fix these features to get a significant boost in accuracy!!",
      "comment_count": 52,
      "views": 0,
      "votes": 62
    },
    {
      "id": 291905,
      "title": "Parquet, and sharing data between notebooks",
      "comment_count": 3,
      "views": 0,
      "votes": 12
    },
    {
      "id": 294843,
      "title": "Basic Recipe for Deep Learning 🔥",
      "comment_count": 5,
      "views": 0,
      "votes": 7
    },
    {
      "id": 295111,
      "title": "CTGAN relevancy, PII anonymization and helping charities",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 295398,
      "title": "Pseudolabeling + Features Engineering",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 294062,
      "title": "Tabular Data Competition Solutions in 2021",
      "comment_count": 21,
      "views": 0,
      "votes": 51
    },
    {
      "id": 294929,
      "title": "How to add covtype.csv to your notebook",
      "comment_count": 12,
      "views": 0,
      "votes": 4
    },
    {
      "id": 295003,
      "title": "Logistic Regression|Neural Networks|XGBoost|Random Forests 📋 🏆",
      "comment_count": 0,
      "views": 0,
      "votes": 9
    },
    {
      "id": 294056,
      "title": "Who are looking forward Kaggle Staff's decision ?",
      "comment_count": 7,
      "views": 0,
      "votes": 19
    },
    {
      "id": 294858,
      "title": "⭐️ Don't worry! Label 5 does not exist in test data!",
      "comment_count": 4,
      "views": 0,
      "votes": 5
    },
    {
      "id": 294875,
      "title": "Synthetic Data Augmentation - CTGAN (Tabular GAN)",
      "comment_count": 4,
      "views": 0,
      "votes": 6
    },
    {
      "id": 293728,
      "title": "Semi-Supervised Learning Papers ",
      "comment_count": 7,
      "views": 0,
      "votes": 16
    },
    {
      "id": 294670,
      "title": "Learning About Denoising AutoEncoders - DAE",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 294808,
      "title": "Exploiting sparsity to reduce memory usage and speed up xgboost training",
      "comment_count": 0,
      "views": 0,
      "votes": 9
    },
    {
      "id": 295167,
      "title": "Multiclass ensemble using predict_proba",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 291844,
      "title": "Reduce the size of your train and test data to model more easily",
      "comment_count": 22,
      "views": 0,
      "votes": 56
    },
    {
      "id": 295458,
      "title": "Exciting competition",
      "comment_count": 2,
      "views": 0,
      "votes": -3
    },
    {
      "id": 294433,
      "title": "Is oversampling reducing accuracy?",
      "comment_count": 9,
      "views": 0,
      "votes": 6
    },
    {
      "id": 293256,
      "title": "Denoising AutoEncoder Pytorch Starter",
      "comment_count": 8,
      "views": 0,
      "votes": 16
    },
    {
      "id": 293768,
      "title": "Yet another score booster!",
      "comment_count": 11,
      "views": 0,
      "votes": 22
    },
    {
      "id": 294579,
      "title": "LightGBM not reproducible for multiclass problem?",
      "comment_count": 13,
      "views": 0,
      "votes": 3
    },
    {
      "id": 292100,
      "title": "Can we trust the feature named Wilderness_Area3  ?",
      "comment_count": 9,
      "views": 0,
      "votes": 17
    },
    {
      "id": 294764,
      "title": "Residual Gates (NN) - Any Success ?",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 294661,
      "title": "Comparing with Original Competition Dataset",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 295019,
      "title": "Feature engineering nonlinear variables",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 293951,
      "title": "Different NN Architectures & Implementation (..and just a thank you!)",
      "comment_count": 7,
      "views": 0,
      "votes": 8
    },
    {
      "id": 294806,
      "title": "TPS DEC21,  Beginner Code Notebook Feedback",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 295078,
      "title": "How to cite the dataset in this competition？",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 294743,
      "title": "Comparing covtype and CTGAN describes",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 294077,
      "title": "When your score doesn't improve..",
      "comment_count": 5,
      "views": 0,
      "votes": 7
    },
    {
      "id": 294625,
      "title": "Elevation is OP",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 295004,
      "title": "Dropping 'Wildness_Area3'?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 293115,
      "title": "Class Ordering and Logistic Regression",
      "comment_count": 7,
      "views": 0,
      "votes": 22
    },
    {
      "id": 293767,
      "title": "The last Tabular Playground Series in 2021!",
      "comment_count": 9,
      "views": 0,
      "votes": 16
    },
    {
      "id": 293509,
      "title": "Downsampling easy data",
      "comment_count": 3,
      "views": 0,
      "votes": 9
    },
    {
      "id": 291871,
      "title": "How to deal with imbalance data? (Thoughts, Ideas & Posts)",
      "comment_count": 38,
      "views": 0,
      "votes": 34
    },
    {
      "id": 294411,
      "title": "Dimensionality Reduction",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 294356,
      "title": "Reduce memory and if statements usage.",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 294186,
      "title": "Does private leaderboard generally follow class distribution of public leaderboard?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 293371,
      "title": "Multiclass Focal Loss for LightGBM",
      "comment_count": 1,
      "views": 0,
      "votes": 11
    },
    {
      "id": 293107,
      "title": "The importance of Sharing and \"Recycling\" code.",
      "comment_count": 7,
      "views": 0,
      "votes": 27
    },
    {
      "id": 293362,
      "title": "We really should not be using the accuracy score",
      "comment_count": 11,
      "views": 0,
      "votes": 29
    },
    {
      "id": 293090,
      "title": "TabNet: The End of Gradient Boosting?",
      "comment_count": 4,
      "views": 0,
      "votes": 10
    },
    {
      "id": 294619,
      "title": "A look at how seed works!!",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 293500,
      "title": "Manhattan Distance of Vertical and Horizontal Hydrology Distances",
      "comment_count": 3,
      "views": 0,
      "votes": 14
    },
    {
      "id": 293504,
      "title": "Handling 'Soil_Type15' and 'Soil_Type7' Features",
      "comment_count": 2,
      "views": 0,
      "votes": 10
    },
    {
      "id": 293350,
      "title": "Singleton array xx cannot be considered a valid collection.",
      "comment_count": 3,
      "views": 0,
      "votes": 11
    },
    {
      "id": 293444,
      "title": "How to implement plurality voting?",
      "comment_count": 3,
      "views": 0,
      "votes": 11
    },
    {
      "id": 293072,
      "title": "Focal loss, a solution at the loss level for imbalanced data",
      "comment_count": 3,
      "views": 0,
      "votes": 23
    },
    {
      "id": 293989,
      "title": "GANs detection",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 291787,
      "title": "Load Data Faster using Feather",
      "comment_count": 3,
      "views": 0,
      "votes": 8
    },
    {
      "id": 293690,
      "title": "How share my new notebook with you?",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 293064,
      "title": "How to deal with class 5 which has only 1 sample",
      "comment_count": 7,
      "views": 0,
      "votes": 10
    },
    {
      "id": 291840,
      "title": "Additional information about Soil Types (from previous competition)",
      "comment_count": 3,
      "views": 0,
      "votes": 19
    },
    {
      "id": 292387,
      "title": "Probing the public leaderboard for rarer classes",
      "comment_count": 4,
      "views": 0,
      "votes": 18
    },
    {
      "id": 291875,
      "title": "single instance of Cover_Type = 5?",
      "comment_count": 9,
      "views": 0,
      "votes": 9
    },
    {
      "id": 292826,
      "title": "Papers With Code: Latest on Deep Learning for Tabular Data",
      "comment_count": 4,
      "views": 0,
      "votes": 11
    },
    {
      "id": 291832,
      "title": "The Covertype dataset, everything you need to know",
      "comment_count": 2,
      "views": 0,
      "votes": 29
    },
    {
      "id": 292793,
      "title": "How to reduce Our Data in terms of rows without losing Accuracy?",
      "comment_count": 6,
      "views": 0,
      "votes": 8
    },
    {
      "id": 292069,
      "title": "TPS December 2021: \"Winter\"",
      "comment_count": 2,
      "views": 0,
      "votes": 16
    },
    {
      "id": 291839,
      "title": "Finally you can try a lot of feature engineering!",
      "comment_count": 0,
      "views": 0,
      "votes": 32
    },
    {
      "id": 291835,
      "title": "Classification: How imbalanced is \"imbalanced\"?",
      "comment_count": 7,
      "views": 0,
      "votes": 15
    },
    {
      "id": 292456,
      "title": "Clustering the data without failing for overfitting specific areas",
      "comment_count": 0,
      "views": 0,
      "votes": 11
    },
    {
      "id": 291774,
      "title": "Looking for a Team Megathread",
      "comment_count": 2,
      "views": 0,
      "votes": 7
    },
    {
      "id": 292526,
      "title": "Metric for leaderboard: categorization accuracy",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 292824,
      "title": "SAINT+: A Transformer-based model for combining self-attention with inter-sample attention over multiple rows",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 292796,
      "title": "Problems with Pytorch and how to solved them",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 291870,
      "title": " Dealing with Imbalanced Datasets",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 291821,
      "title": "PyTorch Series : Folds, Datasets and DataLoaders, nn.LazyLinear, Accuracy metric",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 291819,
      "title": "💥🎉 Most Voted Notebook From Previous Competition ",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 291816,
      "title": "training-on-gpu (CatBoostClassifier - XGBClassifier)",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    }
  ],
  "errors": []
}