{
  "total": 96,
  "topics": [
    {
      "id": 275669,
      "title": "1M rows: ...how to read in only some of the data",
      "comment_count": 32,
      "views": 0,
      "votes": 75
    },
    {
      "id": 284492,
      "title": "✔️ 9th place solution",
      "comment_count": 19,
      "views": 0,
      "votes": 40
    },
    {
      "id": 284594,
      "title": "3rd place solution",
      "comment_count": 5,
      "views": 0,
      "votes": 31
    },
    {
      "id": 284560,
      "title": "4th place solution",
      "comment_count": 2,
      "views": 0,
      "votes": 16
    },
    {
      "id": 283847,
      "title": "What I learned from my first Kaggle Competition",
      "comment_count": 6,
      "views": 0,
      "votes": 21
    },
    {
      "id": 276162,
      "title": "Datatable loads data 6.6x faster than pandas!",
      "comment_count": 8,
      "views": 0,
      "votes": 16
    },
    {
      "id": 284245,
      "title": "Publishing a high scoring notebook",
      "comment_count": 21,
      "views": 0,
      "votes": 7
    },
    {
      "id": 284631,
      "title": "Thoughts on my work on the Oct 2021 Tabular competition",
      "comment_count": 1,
      "views": 0,
      "votes": 9
    },
    {
      "id": 284176,
      "title": "What I learnt from TPS Oct '21 as an absolute beginner",
      "comment_count": 2,
      "views": 0,
      "votes": 12
    },
    {
      "id": 284785,
      "title": "TPS - Oct 2021 - Solutions compilation ",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 283907,
      "title": "What are your moves for the last 3 days of the competition?",
      "comment_count": 5,
      "views": 0,
      "votes": 14
    },
    {
      "id": 284279,
      "title": "[Help] How does scipy.optimise minimize figure out weights for different models in blending",
      "comment_count": 3,
      "views": 0,
      "votes": 6
    },
    {
      "id": 283885,
      "title": "How to save your trained model",
      "comment_count": 5,
      "views": 0,
      "votes": 20
    },
    {
      "id": 284447,
      "title": "I used XGBoost, what more could I have done to improve?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 284452,
      "title": "My Blending Did Better Than My Stacking",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 282785,
      "title": "[TPS Oct 2021] Things worth trying out this month",
      "comment_count": 4,
      "views": 0,
      "votes": 14
    },
    {
      "id": 280661,
      "title": "Ignore some folds in the cross-validation method & MORE",
      "comment_count": 14,
      "views": 0,
      "votes": 28
    },
    {
      "id": 281993,
      "title": "Has anyone tried Boruta-SHAP on LightGBM?",
      "comment_count": 19,
      "views": 0,
      "votes": 4
    },
    {
      "id": 283924,
      "title": "Do people are not interested in working with large dataset? why?",
      "comment_count": 6,
      "views": 0,
      "votes": 2
    },
    {
      "id": 284296,
      "title": "First competition",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 282125,
      "title": "Catboost Optuna search stops after few trials",
      "comment_count": 9,
      "views": 0,
      "votes": 5
    },
    {
      "id": 284071,
      "title": "Don't forget to predict_proba",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 275609,
      "title": "🔥🔥TPS Oct 2021 - 🔥🔥Some great code to look at 🔥🔥 for starters",
      "comment_count": 6,
      "views": 0,
      "votes": 17
    },
    {
      "id": 279277,
      "title": "The need for KMeans",
      "comment_count": 10,
      "views": 0,
      "votes": 17
    },
    {
      "id": 280986,
      "title": "What's your thought in this👇??",
      "comment_count": 12,
      "views": 0,
      "votes": 9
    },
    {
      "id": 280730,
      "title": "Suggestions for improvement",
      "comment_count": 8,
      "views": 0,
      "votes": 3
    },
    {
      "id": 280437,
      "title": "The need for KMeans++",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 280091,
      "title": "It seems that AdaBoost and GradientBoost classifier from sklearn are slow and less effective",
      "comment_count": 6,
      "views": 0,
      "votes": 5
    },
    {
      "id": 280110,
      "title": "My Keras TensorFlow model has AUC always at .5",
      "comment_count": 7,
      "views": 0,
      "votes": 1
    },
    {
      "id": 279197,
      "title": "Is hyperparameter tuning difficult?",
      "comment_count": 10,
      "views": 0,
      "votes": 7
    },
    {
      "id": 280139,
      "title": " No module named 'cudf' who can help?",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 275712,
      "title": "🔥🔥Tutorial compilation for handling larger datasets",
      "comment_count": 7,
      "views": 0,
      "votes": 56
    },
    {
      "id": 277409,
      "title": "[Compilation] Good starter notebooks",
      "comment_count": 8,
      "views": 0,
      "votes": 35
    },
    {
      "id": 278046,
      "title": "Understanding LightGBM thoroughly - A primer",
      "comment_count": 16,
      "views": 0,
      "votes": 8
    },
    {
      "id": 277700,
      "title": "categorical versus continuous features in Octobers Playground",
      "comment_count": 4,
      "views": 0,
      "votes": 13
    },
    {
      "id": 275733,
      "title": "Understanding the Evaluation metric: AUC ( Area Under Curve)",
      "comment_count": 1,
      "views": 0,
      "votes": 12
    },
    {
      "id": 275916,
      "title": "How to diversify your Blend with non Boosting models?",
      "comment_count": 5,
      "views": 0,
      "votes": 11
    },
    {
      "id": 279254,
      "title": "XGBoost Hyperparameter Tuning using Bayesian optimization",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 279879,
      "title": "when  to split test and valid when using using optuna",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 278968,
      "title": "One man army competition !!",
      "comment_count": 1,
      "views": 0,
      "votes": 5
    },
    {
      "id": 278796,
      "title": "Autoviz: Automatically Visualize any Dataset",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 278520,
      "title": "AUC 0.85671 has been crossed !",
      "comment_count": 10,
      "views": 0,
      "votes": 14
    },
    {
      "id": 278712,
      "title": "Synthetic Data Augmentation - CTGAN (Tabular GAN)",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 278541,
      "title": "Guessing the future winning solutions - common tricks from past kaggle tabular competitions",
      "comment_count": 2,
      "views": 0,
      "votes": 9
    },
    {
      "id": 276953,
      "title": "Don't trust built-in feature importances of any model - use SHAP",
      "comment_count": 18,
      "views": 0,
      "votes": 36
    },
    {
      "id": 279009,
      "title": "Is f22 an outlier?",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 278216,
      "title": "[Compilation]Emsembling ideas",
      "comment_count": 4,
      "views": 0,
      "votes": 9
    },
    {
      "id": 277966,
      "title": "Predicting a Biological Response- Lets learn from history ",
      "comment_count": 4,
      "views": 0,
      "votes": 12
    },
    {
      "id": 275837,
      "title": "Load the data on cpu in less that 5 seconds with feather",
      "comment_count": 1,
      "views": 0,
      "votes": 9
    },
    {
      "id": 278396,
      "title": "Summarizing Feature Selection",
      "comment_count": 0,
      "views": 0,
      "votes": 9
    },
    {
      "id": 277659,
      "title": "Stacking on different seeds of base models",
      "comment_count": 9,
      "views": 0,
      "votes": 5
    },
    {
      "id": 277685,
      "title": "Here the steps (code) to move to Google Colab smoothly without downloading datasets",
      "comment_count": 7,
      "views": 0,
      "votes": 11
    },
    {
      "id": 278219,
      "title": "My TPS Ocotber 2021 | Full Dataset",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 275641,
      "title": "Understand Evaluation metric For this Competition - ROC Curve?",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 278758,
      "title": "Something Useful",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 278237,
      "title": "Two points on reduce memory",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 275930,
      "title": "When using GPU to speed up, pay attention to LightGBM",
      "comment_count": 13,
      "views": 0,
      "votes": 25
    },
    {
      "id": 276400,
      "title": "Anyone able to run LGBM on GPU?",
      "comment_count": 16,
      "views": 0,
      "votes": 9
    },
    {
      "id": 276103,
      "title": "A weird idea",
      "comment_count": 16,
      "views": 0,
      "votes": 14
    },
    {
      "id": 277664,
      "title": "Would it make sense to label encode some of the features with distinct peaks?",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 276043,
      "title": "Sum of Binary Features Per Row",
      "comment_count": 10,
      "views": 0,
      "votes": 11
    },
    {
      "id": 278002,
      "title": "Feature selection idea",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 276222,
      "title": "Collection of baseline approaches",
      "comment_count": 6,
      "views": 0,
      "votes": 25
    },
    {
      "id": 277930,
      "title": "Might be helpful",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 275605,
      "title": "feature22 - correlation",
      "comment_count": 18,
      "views": 0,
      "votes": 27
    },
    {
      "id": 277606,
      "title": "Cross Validation Dilemma: KFold vs. StratifiedKFold",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 276601,
      "title": "XGBoost/LightGBM: Do we ever need to tune the learning rate?",
      "comment_count": 13,
      "views": 0,
      "votes": 7
    },
    {
      "id": 277618,
      "title": "Tabnet on Tabular Playground-Oct",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 277366,
      "title": "I've made a lot of improvements but..",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 277227,
      "title": "A new method of stacking?",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 277436,
      "title": "Question on Scaling",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 275704,
      "title": "EDA for huge datasets?",
      "comment_count": 4,
      "views": 0,
      "votes": 5
    },
    {
      "id": 275670,
      "title": "How to speed up Pandas (Article)",
      "comment_count": 10,
      "views": 0,
      "votes": 11
    },
    {
      "id": 276416,
      "title": "Strange Features with Very High Std Deviation!",
      "comment_count": 6,
      "views": 0,
      "votes": 5
    },
    {
      "id": 276712,
      "title": "Adversarial Validation",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 276069,
      "title": "Simple blend: geomean of scaled top 6 notebooks",
      "comment_count": 5,
      "views": 0,
      "votes": 12
    },
    {
      "id": 277289,
      "title": "Importance of roc_auc_score values",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 276737,
      "title": "what is Accelerator warning that Kaggle pops up for GPU?",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 276650,
      "title": "ROC and AUC",
      "comment_count": 1,
      "views": 0,
      "votes": 5
    },
    {
      "id": 276034,
      "title": "Reduce memory usage BUT FASTER",
      "comment_count": 5,
      "views": 0,
      "votes": 18
    },
    {
      "id": 275854,
      "title": "Save memory for running your best models",
      "comment_count": 3,
      "views": 0,
      "votes": 23
    },
    {
      "id": 276053,
      "title": "Search for Categorical Correlation!!",
      "comment_count": 5,
      "views": 0,
      "votes": 10
    },
    {
      "id": 276349,
      "title": "Not able to train ANN model using Kaggle GPU accelerator",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 275909,
      "title": "Feature Engineering Ideas",
      "comment_count": 7,
      "views": 0,
      "votes": 10
    },
    {
      "id": 275892,
      "title": "⁉️🤔🌊 How to tackle limited memory issue for large tabular dataset ?",
      "comment_count": 4,
      "views": 0,
      "votes": 8
    },
    {
      "id": 275663,
      "title": "Getting memory exceeded | reduce memory",
      "comment_count": 9,
      "views": 0,
      "votes": 14
    },
    {
      "id": 275719,
      "title": "[Utility] Multiple file formats for faster data loading",
      "comment_count": 3,
      "views": 0,
      "votes": 6
    },
    {
      "id": 275794,
      "title": "💥Your notebook tried to allocate more memory than is available. It has restarted.",
      "comment_count": 4,
      "views": 0,
      "votes": 9
    },
    {
      "id": 276130,
      "title": "How do I use GPU to speed up the process?",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 275801,
      "title": "Best Single Model Scores. There's time for ensemble!",
      "comment_count": 1,
      "views": 0,
      "votes": 11
    },
    {
      "id": 276368,
      "title": "XGBoost Part 3: Tree Methods",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 275913,
      "title": "What I have learnt participating in the last 7 Tabular Playground series competitions",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 275891,
      "title": "Reference Documentation For All The Features",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 275954,
      "title": "AutoML Starter",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 275781,
      "title": "Optuna Trials keep on failing?",
      "comment_count": 6,
      "views": 0,
      "votes": 4
    },
    {
      "id": 275701,
      "title": "  @vopani Datatable tutorial",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    }
  ],
  "errors": []
}