{
  "total": 111,
  "topics": [
    {
      "id": 284445,
      "title": "Feedback Requested",
      "comment_count": 89,
      "views": 0,
      "votes": 77
    },
    {
      "id": 288064,
      "title": "Courses about Pytorch, others frameworks, and posts about NN in general",
      "comment_count": 29,
      "views": 0,
      "votes": 34
    },
    {
      "id": 311971,
      "title": "Repo for training and tracking models with MLflow",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 285707,
      "title": "Dealing with Noisy Labels",
      "comment_count": 10,
      "views": 0,
      "votes": 11
    },
    {
      "id": 287989,
      "title": "[Productivity Guide] - How to work faster with notebooks ",
      "comment_count": 20,
      "views": 0,
      "votes": 80
    },
    {
      "id": 287984,
      "title": "What is the Logical Steps(sequence) on building deep learning models?",
      "comment_count": 6,
      "views": 0,
      "votes": 11
    },
    {
      "id": 284475,
      "title": "TPS-2021 - Previous Winning solutions",
      "comment_count": 3,
      "views": 0,
      "votes": 16
    },
    {
      "id": 291883,
      "title": "#1 solution",
      "comment_count": 10,
      "views": 0,
      "votes": 21
    },
    {
      "id": 291766,
      "title": "#3 Solution: Don't trust the cv scores",
      "comment_count": 14,
      "views": 0,
      "votes": 27
    },
    {
      "id": 291494,
      "title": "Overfitting tool has just been published for the curious :-)",
      "comment_count": 9,
      "views": 0,
      "votes": 35
    },
    {
      "id": 291091,
      "title": "Is adaptive overfitting occurring on the leaderboard?",
      "comment_count": 12,
      "views": 0,
      "votes": 21
    },
    {
      "id": 291903,
      "title": "#2 solution",
      "comment_count": 2,
      "views": 0,
      "votes": 9
    },
    {
      "id": 291846,
      "title": "#5 solution (and +141 from Public Leaderboard)",
      "comment_count": 2,
      "views": 0,
      "votes": 10
    },
    {
      "id": 291259,
      "title": "By far the best video I've seen about How Neural Networks works.",
      "comment_count": 4,
      "views": 0,
      "votes": 19
    },
    {
      "id": 291463,
      "title": "Data generated by sklearn's make_classification?",
      "comment_count": 1,
      "views": 0,
      "votes": 14
    },
    {
      "id": 291761,
      "title": "Without leaked target, private lb can reach 0.75200",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 291781,
      "title": "Compliments to the winners and looking forward to read about your solutions :-)",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 289595,
      "title": "Difference between NN (Keras) and non-NN models",
      "comment_count": 10,
      "views": 0,
      "votes": 28
    },
    {
      "id": 291003,
      "title": "Before choosing your submission, have a look at the adversarial validation",
      "comment_count": 6,
      "views": 0,
      "votes": 16
    },
    {
      "id": 291317,
      "title": "Ensembling gone?",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 290810,
      "title": "Why cv scores are unreliable in this competition",
      "comment_count": 9,
      "views": 0,
      "votes": 16
    },
    {
      "id": 291073,
      "title": "How to Tune Neural Networks with OPTUNA?",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 291101,
      "title": "Huge gap in validation and submission aucs",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 291132,
      "title": "Inspiration ideas for your next scikit-learn creations",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 290593,
      "title": "How is the submission score calculated?",
      "comment_count": 4,
      "views": 0,
      "votes": 8
    },
    {
      "id": 288221,
      "title": "Guessing the future winning solutions ",
      "comment_count": 11,
      "views": 0,
      "votes": 34
    },
    {
      "id": 290723,
      "title": "I increased my patience, folds, and added a plateau and now my NN overfits",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 291282,
      "title": "TPU and Tensorflow version",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 285092,
      "title": "Triple mastership!",
      "comment_count": 84,
      "views": 0,
      "votes": 83
    },
    {
      "id": 289034,
      "title": "Looking at flip probability between the chunks",
      "comment_count": 23,
      "views": 0,
      "votes": 16
    },
    {
      "id": 289267,
      "title": "There will be a December TPS?",
      "comment_count": 8,
      "views": 0,
      "votes": 26
    },
    {
      "id": 290297,
      "title": "Hard margin SVM by cvxopt solver",
      "comment_count": 0,
      "views": 0,
      "votes": 10
    },
    {
      "id": 290407,
      "title": "How to do different predictions for data on different sides of a linear regression in sklearn?",
      "comment_count": 1,
      "views": 0,
      "votes": 5
    },
    {
      "id": 285772,
      "title": "TPS Nov: Find the magic  😄",
      "comment_count": 2,
      "views": 0,
      "votes": 23
    },
    {
      "id": 290793,
      "title": "Neural Network for this competition",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 290791,
      "title": "Tensorflow vs pytorch",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 289438,
      "title": "Take a break with some pretty plots",
      "comment_count": 4,
      "views": 0,
      "votes": 18
    },
    {
      "id": 289431,
      "title": "Some Activation Functions in ANN 🔥🔥",
      "comment_count": 8,
      "views": 0,
      "votes": 12
    },
    {
      "id": 289619,
      "title": "CV, Overfitting, and shake-down/ups.",
      "comment_count": 2,
      "views": 0,
      "votes": 10
    },
    {
      "id": 290841,
      "title": "Data size reduction",
      "comment_count": 0,
      "views": 0,
      "votes": -2
    },
    {
      "id": 290840,
      "title": "Top 10 model in TPS-Oct",
      "comment_count": 0,
      "views": 0,
      "votes": -2
    },
    {
      "id": 290691,
      "title": "Feature engineering",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 290466,
      "title": "DataOps Dev Engineer ",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 286731,
      "title": "The data is chunked !",
      "comment_count": 25,
      "views": 0,
      "votes": 28
    },
    {
      "id": 288125,
      "title": "Are the leaderboard's >.75 entries overfitting?",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 289807,
      "title": "Want stable CVs?",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 289405,
      "title": "Model stacking ",
      "comment_count": 4,
      "views": 0,
      "votes": 5
    },
    {
      "id": 284784,
      "title": "Why does Neural Network works so well this month?",
      "comment_count": 14,
      "views": 0,
      "votes": 13
    },
    {
      "id": 289317,
      "title": "Permutation importance for each fold???",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 289920,
      "title": "NN: accuracy, recall, F1... ",
      "comment_count": 5,
      "views": 0,
      "votes": 0
    },
    {
      "id": 288433,
      "title": "PyTorch Series : Folds, Datasets and DataLoaders, nn.LazyLinear, AUC metric",
      "comment_count": 2,
      "views": 0,
      "votes": 19
    },
    {
      "id": 288789,
      "title": "Changing target to (0,1,2,3) helps to improve prediction ",
      "comment_count": 4,
      "views": 0,
      "votes": 7
    },
    {
      "id": 285229,
      "title": "[TPS Nov 2021] Things worth trying out this month",
      "comment_count": 14,
      "views": 0,
      "votes": 13
    },
    {
      "id": 289697,
      "title": "Submission range",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 284608,
      "title": "Feature selection may help you (30% reduction of the data)",
      "comment_count": 12,
      "views": 0,
      "votes": 17
    },
    {
      "id": 285503,
      "title": "Mislabeled samples",
      "comment_count": 121,
      "views": 0,
      "votes": 47
    },
    {
      "id": 287707,
      "title": "🔥🔥 Differential Privacy -- Is Anonymization of features sufficient ??",
      "comment_count": 4,
      "views": 0,
      "votes": 17
    },
    {
      "id": 287152,
      "title": "Create your Pytorch trainer with Tez ",
      "comment_count": 3,
      "views": 0,
      "votes": 13
    },
    {
      "id": 287047,
      "title": "Leaked train.csv with original labels",
      "comment_count": 6,
      "views": 0,
      "votes": 16
    },
    {
      "id": 288089,
      "title": "Oscillating loss and AUC for NN",
      "comment_count": 12,
      "views": 0,
      "votes": 6
    },
    {
      "id": 287238,
      "title": "Tips on getting stated with Keras",
      "comment_count": 9,
      "views": 0,
      "votes": 9
    },
    {
      "id": 286814,
      "title": "any ideas to improve the model performance?",
      "comment_count": 11,
      "views": 0,
      "votes": 2
    },
    {
      "id": 287429,
      "title": "Score Comparison by Model Type",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 284461,
      "title": "💫💥 Good Starter References for TPS - Nov 2021 💫💥 ",
      "comment_count": 2,
      "views": 0,
      "votes": 12
    },
    {
      "id": 287039,
      "title": "TPS Nov AutoViML Deep Learning model in less than 5 mins - 0.7446 score",
      "comment_count": 2,
      "views": 0,
      "votes": 14
    },
    {
      "id": 286579,
      "title": "What can be the secret recipe of Laurent Pourchot in every TPS?",
      "comment_count": 19,
      "views": 0,
      "votes": 10
    },
    {
      "id": 285700,
      "title": "Is he a thief?",
      "comment_count": 25,
      "views": 0,
      "votes": 38
    },
    {
      "id": 284777,
      "title": "A sure bet: denoising auto-encoders",
      "comment_count": 26,
      "views": 0,
      "votes": 14
    },
    {
      "id": 285738,
      "title": "Learning >= Winning",
      "comment_count": 5,
      "views": 0,
      "votes": 16
    },
    {
      "id": 286106,
      "title": "Picking neural or boosting",
      "comment_count": 8,
      "views": 0,
      "votes": 14
    },
    {
      "id": 287131,
      "title": "The Data is Skwed",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 286538,
      "title": "Stuck in competition?",
      "comment_count": 8,
      "views": 0,
      "votes": 8
    },
    {
      "id": 286543,
      "title": "They did not randomly flip 25 % of the labels",
      "comment_count": 8,
      "views": 0,
      "votes": 14
    },
    {
      "id": 286610,
      "title": "Tuning your neural network architecture for a better score",
      "comment_count": 0,
      "views": 0,
      "votes": 20
    },
    {
      "id": 286475,
      "title": "Auto Submitting from Google Colab via Kaggle API",
      "comment_count": 5,
      "views": 0,
      "votes": 13
    },
    {
      "id": 284835,
      "title": "Tips and tricks for training better Neural Networks",
      "comment_count": 1,
      "views": 0,
      "votes": 25
    },
    {
      "id": 286539,
      "title": "Bi Modal Features",
      "comment_count": 2,
      "views": 0,
      "votes": 7
    },
    {
      "id": 286276,
      "title": "[Up-to-date Summary] We've hit the quarter mark! these are the stories up to this point!",
      "comment_count": 0,
      "views": 0,
      "votes": 19
    },
    {
      "id": 286879,
      "title": "How to Excel in this competition",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 286763,
      "title": "Help needed with an error",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 285355,
      "title": "Don't be tempted to the leaderboard (avoid overfitt)",
      "comment_count": 26,
      "views": 0,
      "votes": 24
    },
    {
      "id": 286113,
      "title": "CV Average Fold Score > Test Score",
      "comment_count": 8,
      "views": 0,
      "votes": 7
    },
    {
      "id": 285006,
      "title": "Any good advice for beginners",
      "comment_count": 21,
      "views": 0,
      "votes": 22
    },
    {
      "id": 285800,
      "title": "TPS Nov 21: Good Starter Notebooks",
      "comment_count": 1,
      "views": 0,
      "votes": 21
    },
    {
      "id": 284775,
      "title": "Building deep neural networks for tabular data with TensorFlow 2.x",
      "comment_count": 6,
      "views": 0,
      "votes": 29
    },
    {
      "id": 285891,
      "title": "DNN feature importance",
      "comment_count": 3,
      "views": 0,
      "votes": 13
    },
    {
      "id": 285626,
      "title": "Applying PCA",
      "comment_count": 9,
      "views": 0,
      "votes": 6
    },
    {
      "id": 285826,
      "title": "Discover wich feature overfitts more (Article)",
      "comment_count": 2,
      "views": 0,
      "votes": 11
    },
    {
      "id": 284850,
      "title": "Use label-smoothing if you want some free AUC ",
      "comment_count": 4,
      "views": 0,
      "votes": 19
    },
    {
      "id": 285876,
      "title": "What machine one should have to try to participate in this competition ",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 285564,
      "title": "[Tutorial] - The Last Feature Selection Guide You’ll Ever Need",
      "comment_count": 3,
      "views": 0,
      "votes": 14
    },
    {
      "id": 285280,
      "title": "Feature Ranking w/ UMAP",
      "comment_count": 3,
      "views": 0,
      "votes": 11
    },
    {
      "id": 285625,
      "title": "Saving Version Time",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 285550,
      "title": "I submitted all the AutoML frameworks so you don't have to 🔥",
      "comment_count": 1,
      "views": 0,
      "votes": 10
    },
    {
      "id": 286293,
      "title": "Things I've noticed and tried",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 285417,
      "title": "Torch, Keras, LightGBM, CatBoost, XGBoost for R Lovers",
      "comment_count": 1,
      "views": 0,
      "votes": 7
    },
    {
      "id": 284691,
      "title": "Seeking Teammate",
      "comment_count": 4,
      "views": 0,
      "votes": 6
    },
    {
      "id": 285508,
      "title": "Using LDA for dimensionality reduction and classification",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 284507,
      "title": "[Guide] - How to trick AUC into scoring your stupid models higher 👌",
      "comment_count": 3,
      "views": 0,
      "votes": 30
    },
    {
      "id": 284618,
      "title": "Need advice on managing notebook versions",
      "comment_count": 7,
      "views": 0,
      "votes": 6
    },
    {
      "id": 285596,
      "title": "Can I apply TTA in this competition?",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 284577,
      "title": "Feature Importances and XAI",
      "comment_count": 2,
      "views": 0,
      "votes": 11
    },
    {
      "id": 284527,
      "title": "Evaluation Metric",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 284692,
      "title": "XGBoost and Scaling?",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 284495,
      "title": "[Compilation] Good starter notebooks from previous TPS",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 284500,
      "title": "Understanding AUC metric",
      "comment_count": 1,
      "views": 0,
      "votes": 11
    },
    {
      "id": 284540,
      "title": "Advanced EDA with UMAP/hdbscan",
      "comment_count": 2,
      "views": 0,
      "votes": 7
    },
    {
      "id": 285439,
      "title": "Memory Reduction Function",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 284636,
      "title": "Rank Gauss dataset",
      "comment_count": 1,
      "views": 0,
      "votes": 5
    },
    {
      "id": 284454,
      "title": "Someone got 0.99999 already",
      "comment_count": 10,
      "views": 0,
      "votes": 4
    },
    {
      "id": 284455,
      "title": "Labels have been updated",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    }
  ],
  "errors": []
}