{
  "total": 370,
  "topics": [
    {
      "id": 56561,
      "title": "Thank you from TalkingData!",
      "comment_count": 4,
      "views": 0,
      "votes": 18
    },
    {
      "id": 51142,
      "title": "Welcome",
      "comment_count": 65,
      "views": 0,
      "votes": 27
    },
    {
      "id": 52658,
      "title": "test_supplement.csv",
      "comment_count": 18,
      "views": 0,
      "votes": 43
    },
    {
      "id": 51962,
      "title": "March Kernels Awards!",
      "comment_count": 15,
      "views": 0,
      "votes": 13
    },
    {
      "id": 51210,
      "title": "Please Check / Re-download test.csv",
      "comment_count": 12,
      "views": 0,
      "votes": 10
    },
    {
      "id": 51176,
      "title": "Welcome from TalkingData!",
      "comment_count": 5,
      "views": 0,
      "votes": 12
    },
    {
      "id": 56105,
      "title": "How to use more features to training with dark numpy magic...",
      "comment_count": 12,
      "views": 0,
      "votes": 87
    },
    {
      "id": 53634,
      "title": "[Updated] Single Best Model Scores : Motivation : Feature Engineering..!!",
      "comment_count": 395,
      "views": 0,
      "votes": 67
    },
    {
      "id": 52263,
      "title": "Can we run multiple kernels at the same time?",
      "comment_count": 8,
      "views": 0,
      "votes": 8
    },
    {
      "id": 54092,
      "title": "Probability calibration",
      "comment_count": 7,
      "views": 0,
      "votes": 5
    },
    {
      "id": 56262,
      "title": "My brief summary,a mainly NN based solution(3th)",
      "comment_count": 108,
      "views": 0,
      "votes": 265
    },
    {
      "id": 51204,
      "title": "IP,Device & OS combination",
      "comment_count": 8,
      "views": 0,
      "votes": 11
    },
    {
      "id": 157481,
      "title": "Your notebook tried to allocate more memory than is available. It has restarted.",
      "comment_count": 6,
      "views": 0,
      "votes": 0
    },
    {
      "id": 56475,
      "title": "1st place solution",
      "comment_count": 95,
      "views": 0,
      "votes": 270
    },
    {
      "id": 56182,
      "title": "We need to talk about kernels (again...)",
      "comment_count": 200,
      "views": 0,
      "votes": 126
    },
    {
      "id": 54948,
      "title": "Load 240MM rows in... 8 seconds",
      "comment_count": 25,
      "views": 0,
      "votes": 46
    },
    {
      "id": 56283,
      "title": "Solution #6 overview",
      "comment_count": 127,
      "views": 0,
      "votes": 240
    },
    {
      "id": 56497,
      "title": "libFM in Keras",
      "comment_count": 29,
      "views": 0,
      "votes": 78
    },
    {
      "id": 53191,
      "title": "Installing R LightGBM",
      "comment_count": 5,
      "views": 0,
      "votes": 0
    },
    {
      "id": 56014,
      "title": "How to access data in gcloud bucket using jupyter notebook",
      "comment_count": 33,
      "views": 0,
      "votes": 4
    },
    {
      "id": 52936,
      "title": "Data Distribution in Train and Test",
      "comment_count": 9,
      "views": 0,
      "votes": 3
    },
    {
      "id": 53696,
      "title": "Calculating class weights",
      "comment_count": 26,
      "views": 0,
      "votes": 35
    },
    {
      "id": 51544,
      "title": "How to resolve the imbalanced data",
      "comment_count": 12,
      "views": 0,
      "votes": 3
    },
    {
      "id": 55107,
      "title": "Dataset size, memory limitations and building a balanced training set",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 56545,
      "title": "Code Sharing, 4th Place",
      "comment_count": 8,
      "views": 0,
      "votes": 44
    },
    {
      "id": 79601,
      "title": "How would they use this result to blacklist ip address as this predicts by clicks not by ips.",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 52777,
      "title": "How to submit from kernel",
      "comment_count": 19,
      "views": 0,
      "votes": 2
    },
    {
      "id": 53753,
      "title": "Is GPU processing now available on Kaggle Kernels?",
      "comment_count": 30,
      "views": 0,
      "votes": 6
    },
    {
      "id": 56333,
      "title": "Hints on 13th place solution.. and finally Grandmaster!",
      "comment_count": 14,
      "views": 0,
      "votes": 43
    },
    {
      "id": 53773,
      "title": "Lightgbm: prevent RAM spike (explode) at the init training",
      "comment_count": 26,
      "views": 0,
      "votes": 66
    },
    {
      "id": 54016,
      "title": "Installing LightGBM GPU on Kaggle Kernels",
      "comment_count": 4,
      "views": 0,
      "votes": 8
    },
    {
      "id": 64988,
      "title": "Notes from my submission",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 56243,
      "title": "4th place (brief) tips",
      "comment_count": 46,
      "views": 0,
      "votes": 136
    },
    {
      "id": 59706,
      "title": "I have got 0.96 locally on validation, but LB only 0.89",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 56217,
      "title": "Now Best way is to recover LB 12hours ago",
      "comment_count": 9,
      "views": 0,
      "votes": 6
    },
    {
      "id": 58415,
      "title": "100% Correlation",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 57971,
      "title": "84th Place Solution with Open Source Code",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 56271,
      "title": "Let us talk about data size",
      "comment_count": 9,
      "views": 0,
      "votes": 15
    },
    {
      "id": 52861,
      "title": "Anyone else having trouble with timeouts?",
      "comment_count": 7,
      "views": 0,
      "votes": null
    },
    {
      "id": 56328,
      "title": "[2nd Place Solution] from PPP",
      "comment_count": 60,
      "views": 0,
      "votes": 114
    },
    {
      "id": 56524,
      "title": "Share Your Public LB Journey",
      "comment_count": 37,
      "views": 0,
      "votes": 17
    },
    {
      "id": 56406,
      "title": "5th place solution (0.9836 with pure feature engineering)",
      "comment_count": 29,
      "views": 0,
      "votes": 57
    },
    {
      "id": 56777,
      "title": "My first kaggle competition journey and what I have learned from winning teams",
      "comment_count": 5,
      "views": 0,
      "votes": 13
    },
    {
      "id": 56250,
      "title": "11th place features",
      "comment_count": 35,
      "views": 0,
      "votes": 98
    },
    {
      "id": 56423,
      "title": "Will it Blend?",
      "comment_count": 5,
      "views": 0,
      "votes": 7
    },
    {
      "id": 56420,
      "title": "East Asia Rocks",
      "comment_count": 53,
      "views": 0,
      "votes": 11
    },
    {
      "id": 56481,
      "title": "Great solution sharing",
      "comment_count": 5,
      "views": 0,
      "votes": 35
    },
    {
      "id": 56241,
      "title": "130 copied submissions in Silver and Bronze",
      "comment_count": 46,
      "views": 0,
      "votes": 53
    },
    {
      "id": 56970,
      "title": "new m",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 56268,
      "title": "Solution to Duplicate Problem by Reverse Engineering (0.0005 Boost)",
      "comment_count": 19,
      "views": 0,
      "votes": 81
    },
    {
      "id": 56272,
      "title": "Small trick to iterate much faster... ",
      "comment_count": 15,
      "views": 0,
      "votes": 27
    },
    {
      "id": 56571,
      "title": "22nd Place Overview",
      "comment_count": 4,
      "views": 0,
      "votes": 15
    },
    {
      "id": 56429,
      "title": "Solution #50 Story - One Week Hackathon ",
      "comment_count": 8,
      "views": 0,
      "votes": 27
    },
    {
      "id": 56319,
      "title": "5th place story (pocket)",
      "comment_count": 4,
      "views": 0,
      "votes": 51
    },
    {
      "id": 56237,
      "title": "Happy and exhausted",
      "comment_count": 40,
      "views": 0,
      "votes": 38
    },
    {
      "id": 56279,
      "title": "9th place",
      "comment_count": 19,
      "views": 0,
      "votes": 49
    },
    {
      "id": 56325,
      "title": "8'th place solution. [ods.ai] blenders in game )",
      "comment_count": 7,
      "views": 0,
      "votes": 45
    },
    {
      "id": 56807,
      "title": "how to accelerate rolling with groupby?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 56304,
      "title": "34th place, thoughts on CV and Target Encoding in R",
      "comment_count": 11,
      "views": 0,
      "votes": 23
    },
    {
      "id": 56245,
      "title": "Kaggle Toxic Submission Classification Challenge is All you Need",
      "comment_count": 12,
      "views": 0,
      "votes": 21
    },
    {
      "id": 56339,
      "title": "The Biggest Failure",
      "comment_count": 9,
      "views": 0,
      "votes": 28
    },
    {
      "id": 56259,
      "title": "560th Solution :)",
      "comment_count": 17,
      "views": 0,
      "votes": 41
    },
    {
      "id": 56205,
      "title": "Does Anyone Have a 0.9835 Kernel?",
      "comment_count": 55,
      "views": 0,
      "votes": 35
    },
    {
      "id": 56368,
      "title": "28th place, A 0.0006x boosting trend feature and non-overfitting target encoding (attributed rates)",
      "comment_count": 21,
      "views": 0,
      "votes": 26
    },
    {
      "id": 56282,
      "title": "Best LibFFM",
      "comment_count": 8,
      "views": 0,
      "votes": 27
    },
    {
      "id": 56317,
      "title": "[late sub 0.9828771]How Boosting from 0.9671 to 0.9823 my first competition on kaggle",
      "comment_count": 6,
      "views": 0,
      "votes": 21
    },
    {
      "id": 56031,
      "title": "Very Disappointed ",
      "comment_count": 19,
      "views": 0,
      "votes": 23
    },
    {
      "id": 56706,
      "title": "Did anyone use Dask or Ray?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 56052,
      "title": "Vote of Thanks and Lessons Learnt",
      "comment_count": 37,
      "views": 0,
      "votes": 44
    },
    {
      "id": 56422,
      "title": "18th Place Brief Solution Overview",
      "comment_count": 3,
      "views": 0,
      "votes": 18
    },
    {
      "id": 56218,
      "title": "Serenity now",
      "comment_count": 8,
      "views": 0,
      "votes": 6
    },
    {
      "id": 56329,
      "title": "Leaderboard Progression Animation",
      "comment_count": 5,
      "views": 0,
      "votes": 19
    },
    {
      "id": 56252,
      "title": "Number of Copied Submissions",
      "comment_count": 0,
      "views": 0,
      "votes": 21
    },
    {
      "id": 56231,
      "title": "Will Dirk's submission lead to disqualification",
      "comment_count": 25,
      "views": 0,
      "votes": 8
    },
    {
      "id": 56201,
      "title": "Is there any way that .9811 csv file can be traced and ignored ?",
      "comment_count": 16,
      "views": 0,
      "votes": 16
    },
    {
      "id": 56349,
      "title": "Solution #55 up on github",
      "comment_count": 1,
      "views": 0,
      "votes": 13
    },
    {
      "id": 56321,
      "title": "How to scavenge kernels",
      "comment_count": 6,
      "views": 0,
      "votes": 2
    },
    {
      "id": 56270,
      "title": "Thanks for this nice competition",
      "comment_count": 2,
      "views": 0,
      "votes": 15
    },
    {
      "id": 56120,
      "title": "How much gain do you get from ensemble model？",
      "comment_count": 24,
      "views": 0,
      "votes": 8
    },
    {
      "id": 56207,
      "title": "0.9813 submission attached",
      "comment_count": 15,
      "views": 0,
      "votes": 13
    },
    {
      "id": 56158,
      "title": "LightGBM Faster Training",
      "comment_count": 6,
      "views": 0,
      "votes": 16
    },
    {
      "id": 56190,
      "title": "good night ,time to sleep",
      "comment_count": 51,
      "views": 0,
      "votes": 3
    },
    {
      "id": 56309,
      "title": "Did anyone get a better result with pseudo-labeling? ",
      "comment_count": 8,
      "views": 0,
      "votes": 6
    },
    {
      "id": 56404,
      "title": "191st Place Solution Or more like lessons learnt....",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 56236,
      "title": "Congrats to The Winners!",
      "comment_count": 2,
      "views": 0,
      "votes": 9
    },
    {
      "id": 56435,
      "title": "multiple filters in CNN gives 0.0004 boost to 0.9780 ",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 56324,
      "title": "Dart was a way better in this competition",
      "comment_count": 12,
      "views": 0,
      "votes": 2
    },
    {
      "id": 56208,
      "title": "This competition is about RAM access",
      "comment_count": 12,
      "views": 0,
      "votes": 1
    },
    {
      "id": 56421,
      "title": "Some minds",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 56233,
      "title": "why someone  do something no-harmonious",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 56312,
      "title": "\"Kaggle-style\" compatible ideas for the (not new) kernels issue?",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 56443,
      "title": "Adversarial Validation within Test Data Makes Sense ?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 56235,
      "title": "Congratulations Winners!!!",
      "comment_count": 7,
      "views": 0,
      "votes": 3
    },
    {
      "id": 55677,
      "title": "Maybe the shakeup is caused from the duplicated data..",
      "comment_count": 9,
      "views": 0,
      "votes": 36
    },
    {
      "id": 56248,
      "title": "Best way to deal with the situation (everyone happy)",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 55601,
      "title": "Some tips to save time and RAM",
      "comment_count": 17,
      "views": 0,
      "votes": 27
    },
    {
      "id": 56232,
      "title": "3 minutes.. good luck for everyone ))",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 56246,
      "title": "Hope kaggle team could consider kernel usage...",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 56280,
      "title": "Technology stack",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 56212,
      "title": "Question about full data set training and validation ",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 56133,
      "title": "Single Best Features",
      "comment_count": 17,
      "views": 0,
      "votes": 0
    },
    {
      "id": 56251,
      "title": "anyone have the 09811 lightgbm kernel?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 56234,
      "title": "Stream not readable",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 56065,
      "title": "Question : How to create Nextclick features in R?",
      "comment_count": 11,
      "views": 0,
      "votes": 1
    },
    {
      "id": 56099,
      "title": "ip%27 leak??? seriously???",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 56114,
      "title": "The remote server returned an error: (404) Not Found.",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 56056,
      "title": "Detailed descriptions of lightgbm parameters",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 56058,
      "title": "Problem",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 56101,
      "title": "Chunk Learning vs One Time Training",
      "comment_count": 2,
      "views": 0,
      "votes": 0
    },
    {
      "id": 56053,
      "title": "Next_click",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 55581,
      "title": "saving/loading feature from local disk",
      "comment_count": 19,
      "views": 0,
      "votes": 5
    },
    {
      "id": 55996,
      "title": "How long will we wait for the final results?",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 55325,
      "title": "Reducing Lightgbm RAM spike using 1 difference",
      "comment_count": 21,
      "views": 0,
      "votes": 36
    },
    {
      "id": 56116,
      "title": "What about a technical topic à la fin",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 55788,
      "title": "Mind Blown",
      "comment_count": 5,
      "views": 0,
      "votes": 10
    },
    {
      "id": 55475,
      "title": "TalkingData LB0.9786",
      "comment_count": 28,
      "views": 0,
      "votes": 15
    },
    {
      "id": 55954,
      "title": "Any advice for hyperparamter optimization when increasing dataset",
      "comment_count": 8,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55771,
      "title": "Shake up is coming?",
      "comment_count": 6,
      "views": 0,
      "votes": 8
    },
    {
      "id": 56068,
      "title": "Offtopic, optimization question",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 55449,
      "title": "LGB using R. Where did I go wrong?",
      "comment_count": 28,
      "views": 0,
      "votes": 2
    },
    {
      "id": 56019,
      "title": "Using parameter tuning but score 0.006 drop. what's problem?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55980,
      "title": "using day 6 as cv?",
      "comment_count": 6,
      "views": 0,
      "votes": 2
    },
    {
      "id": 55821,
      "title": "Choosing entries for final submission",
      "comment_count": 17,
      "views": 0,
      "votes": 2
    },
    {
      "id": 55984,
      "title": "how to fix great gap between local val score and LB",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55928,
      "title": "Data Download Issue",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 55626,
      "title": "Does hyperparameter tuning in lightgbm really helpful in this competition?",
      "comment_count": 10,
      "views": 0,
      "votes": 11
    },
    {
      "id": 54171,
      "title": "Best Single Model Kernel Score?",
      "comment_count": 91,
      "views": 0,
      "votes": 24
    },
    {
      "id": 55858,
      "title": "Submission spends long time to upload a file",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 55932,
      "title": "Is a Single-Model more reliable to evade shake-up? ",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 55884,
      "title": "private leaderboard test data what's time released?",
      "comment_count": 13,
      "views": 0,
      "votes": 0
    },
    {
      "id": 55666,
      "title": "how to creat histogram feature?",
      "comment_count": 7,
      "views": 0,
      "votes": 4
    },
    {
      "id": 55030,
      "title": "28 New Features - 0.9803 LB Score [Updated - v5]",
      "comment_count": 40,
      "views": 0,
      "votes": 18
    },
    {
      "id": 55919,
      "title": "LightGBM stops too early",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 54400,
      "title": "Kernels Schmernels",
      "comment_count": 27,
      "views": 0,
      "votes": 15
    },
    {
      "id": 55563,
      "title": "How to private share but comply with rules",
      "comment_count": 7,
      "views": 0,
      "votes": 15
    },
    {
      "id": 55769,
      "title": "Is there anyone confident ?",
      "comment_count": 9,
      "views": 0,
      "votes": 0
    },
    {
      "id": 55768,
      "title": "Do many kernels share memory usage?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 55766,
      "title": "New Features -> New Hyperparameters",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 55685,
      "title": "Lightgbm categorical features transformation",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55628,
      "title": "Submission Error ",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55593,
      "title": "Using SQL For Complex GROUP BY Features",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55304,
      "title": "open solution (late start) [0.9567]",
      "comment_count": 7,
      "views": 0,
      "votes": 5
    },
    {
      "id": 55338,
      "title": "Any success on using LightGBM or XGboost with GPU acceleration?",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54929,
      "title": "any success with non tree based models?",
      "comment_count": 18,
      "views": 0,
      "votes": 17
    },
    {
      "id": 55166,
      "title": "New Random Seed Increased LB Score by 0.0004. Can I Use it?",
      "comment_count": 19,
      "views": 0,
      "votes": 3
    },
    {
      "id": 54765,
      "title": "Business Context of Click Fraud - Chinese Click Farm",
      "comment_count": 18,
      "views": 0,
      "votes": 73
    },
    {
      "id": 53251,
      "title": "Advice to newbies",
      "comment_count": 54,
      "views": 0,
      "votes": 74
    },
    {
      "id": 55346,
      "title": "In real scenario, these features wouldn't be used",
      "comment_count": 14,
      "views": 0,
      "votes": 4
    },
    {
      "id": 55523,
      "title": "Reason behind Notebook gives better result? LB: 0.9786",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55147,
      "title": "attributed frequencies overfitting",
      "comment_count": 25,
      "views": 0,
      "votes": 11
    },
    {
      "id": 55498,
      "title": "the val score is always higher than train score in training time",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 55054,
      "title": "Business Context of Click Fraud - Chinese Click Farm - V2",
      "comment_count": 3,
      "views": 0,
      "votes": 26
    },
    {
      "id": 55227,
      "title": "weird variable importance in lightGBM",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 52779,
      "title": "NO MORE BLENDING IN KERNALS PLEASE",
      "comment_count": 57,
      "views": 0,
      "votes": 80
    },
    {
      "id": 55398,
      "title": "LightGBM Bosses",
      "comment_count": 6,
      "views": 0,
      "votes": 2
    },
    {
      "id": 55514,
      "title": "Why always Evaluation Exception: Stream was not readable",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 55281,
      "title": "[for fun]What were kagglers doing in the 16th century?",
      "comment_count": 5,
      "views": 0,
      "votes": 3
    },
    {
      "id": 53378,
      "title": "Test and test supplement click ids are not the same",
      "comment_count": 35,
      "views": 0,
      "votes": 17
    },
    {
      "id": 55164,
      "title": "how to use stacking？",
      "comment_count": 10,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55473,
      "title": "Multiple Validation sets LightGBM",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 55379,
      "title": "Save more Memory in Pandas",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 55438,
      "title": "How to tackle memory spikes while merging/feature generation?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55435,
      "title": "Do you use parallel processing ?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 55410,
      "title": "Why my local CV is a bit higher than locan train-set CV?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 55193,
      "title": "Blends and overfitting",
      "comment_count": 13,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54586,
      "title": "Tweak to libffm to get AUC scores",
      "comment_count": 9,
      "views": 0,
      "votes": 12
    },
    {
      "id": 55255,
      "title": "Any successful Neural Net Arch's?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 55277,
      "title": "Is there a function which returns memory in use?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55238,
      "title": "What do these two warnings mean?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 55129,
      "title": "What is the kernel execution time constraint?",
      "comment_count": 9,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55121,
      "title": "This competition is too memory demanding",
      "comment_count": 17,
      "views": 0,
      "votes": 2
    },
    {
      "id": 55111,
      "title": "LGBM Bayesian Optimization",
      "comment_count": 8,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54968,
      "title": "May i do feature engineering on train+test dataset same time ?",
      "comment_count": 7,
      "views": 0,
      "votes": 3
    },
    {
      "id": 55010,
      "title": "What's with user 5314:72:1:13:101, they download 84% of the time",
      "comment_count": 7,
      "views": 0,
      "votes": 1
    },
    {
      "id": 55104,
      "title": "Chinese Click Farm Video",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 55196,
      "title": "Noob remark about training data",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 54908,
      "title": "how to deal with the slow uploading problem?",
      "comment_count": 17,
      "views": 0,
      "votes": null
    },
    {
      "id": 54962,
      "title": "How come single IP has more than one device and OS ?",
      "comment_count": 11,
      "views": 0,
      "votes": 0
    },
    {
      "id": 55051,
      "title": "Has anyone encountered this? Evaluation Exception: Stream was not readable.",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 54620,
      "title": "Lack of model diversity in this competition?",
      "comment_count": 7,
      "views": 0,
      "votes": 8
    },
    {
      "id": 55004,
      "title": "Quick question about helping me to submit my score",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 54723,
      "title": "Accidental interesting finding about validation",
      "comment_count": 6,
      "views": 0,
      "votes": 5
    },
    {
      "id": 54673,
      "title": "Importance of CV - LB constant?",
      "comment_count": 10,
      "views": 0,
      "votes": 9
    },
    {
      "id": 54919,
      "title": "LGBM Tuning Question - Min_child_samples",
      "comment_count": 8,
      "views": 0,
      "votes": null
    },
    {
      "id": 54939,
      "title": "Is click time already sorted?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54648,
      "title": "Why don't we use MAC address to identify the fraud clicks?",
      "comment_count": 9,
      "views": 0,
      "votes": 6
    },
    {
      "id": 54633,
      "title": "Anybody using laptop for training?",
      "comment_count": 8,
      "views": 0,
      "votes": 3
    },
    {
      "id": 54874,
      "title": "TalkingData LGBM ~Public LB 0.9659",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 53450,
      "title": "C++ Time Delta Code",
      "comment_count": 19,
      "views": 0,
      "votes": 23
    },
    {
      "id": 54824,
      "title": "Do we need to upload the code and model? ",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 54825,
      "title": "Solicit advice or code review for out-of-core sklearn RF (ExtraTree)",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 51347,
      "title": "Shrink the data from command line",
      "comment_count": 14,
      "views": 0,
      "votes": 76
    },
    {
      "id": 54803,
      "title": "Questions about attributed_time, need suggestions please.",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 52782,
      "title": "Is this a good competition for pseudo-labeling?",
      "comment_count": 8,
      "views": 0,
      "votes": 6
    },
    {
      "id": 54367,
      "title": "Questions to all the deep learning pros",
      "comment_count": 13,
      "views": 0,
      "votes": 9
    },
    {
      "id": 54686,
      "title": "Online Algorithms in R",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 54521,
      "title": "Is the Leaderboard Difference Meaningful?",
      "comment_count": 6,
      "views": 0,
      "votes": 4
    },
    {
      "id": 54539,
      "title": "How do you choose a logistic regularization parameter?",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 54229,
      "title": "Duplicates",
      "comment_count": 13,
      "views": 0,
      "votes": 13
    },
    {
      "id": 53361,
      "title": "Time Deltas",
      "comment_count": 38,
      "views": 0,
      "votes": 18
    },
    {
      "id": 54284,
      "title": "Do not specialize on n:n rows",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54626,
      "title": "Problem with Notebook kernel",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 54531,
      "title": "test & supplement index mapping",
      "comment_count": 3,
      "views": 0,
      "votes": 6
    },
    {
      "id": 54169,
      "title": "Quirky server",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 54526,
      "title": "Feature Importance Technique",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 54184,
      "title": "questions about feature selection",
      "comment_count": 9,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54298,
      "title": "Submission waiting for a long time",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54483,
      "title": "How to do bagging/blending properly?",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 53886,
      "title": "App Id",
      "comment_count": 19,
      "views": 0,
      "votes": 7
    },
    {
      "id": 54605,
      "title": "The importance of test_sup",
      "comment_count": 2,
      "views": 0,
      "votes": -5
    },
    {
      "id": 54516,
      "title": "Duplicated rows in train and test",
      "comment_count": 1,
      "views": 0,
      "votes": 0
    },
    {
      "id": 53246,
      "title": "features to use",
      "comment_count": 12,
      "views": 0,
      "votes": 1
    },
    {
      "id": 51454,
      "title": "Load the whole dataset in few minutes!",
      "comment_count": 23,
      "views": 0,
      "votes": 11
    },
    {
      "id": 52739,
      "title": "Let's save bandwidth!",
      "comment_count": 24,
      "views": 0,
      "votes": 17
    },
    {
      "id": 54468,
      "title": "anyone in Washington D.C area open for meeting up?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 54224,
      "title": "Why getting error when updating the submission?",
      "comment_count": 12,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54430,
      "title": "Batch Size for training of dataset",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 54364,
      "title": "Why not use auPRC as an evaluation method?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53987,
      "title": "What's your estimated private lb score vs your public lb score?",
      "comment_count": 10,
      "views": 0,
      "votes": 5
    },
    {
      "id": 53141,
      "title": "In Defense of Public Blending: Seek Alpha, Not Total Return",
      "comment_count": 36,
      "views": 0,
      "votes": 16
    },
    {
      "id": 54193,
      "title": "SQL to find click_id from test inside test_sup",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53595,
      "title": "Model Leakage",
      "comment_count": 20,
      "views": 0,
      "votes": 12
    },
    {
      "id": 54160,
      "title": "Evaluation Exception: Submission must have 18790469 rows",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 54165,
      "title": "train.csv.zip Error Unzipping",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 53668,
      "title": "Be careful - Ids are not sorted",
      "comment_count": 24,
      "views": 0,
      "votes": 29
    },
    {
      "id": 54112,
      "title": "Memory Error",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54006,
      "title": "Wish we could submit 10 Predictions/Day",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 53964,
      "title": "Can not load train.csv on kaggle kernel",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 54072,
      "title": "How to groupby more than one column with minimum RAM",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 54053,
      "title": "Trouble uploading predictions",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 53353,
      "title": "A problem with running kernels on Kaggle",
      "comment_count": 18,
      "views": 0,
      "votes": 14
    },
    {
      "id": 54030,
      "title": "Analysing Full Dataset with Apache Spark (local mode or not)",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53972,
      "title": "Kernel : Failed. Exited with code 137.",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 53993,
      "title": "Is the overlap useful?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 53989,
      "title": "Sample Training Based on the Testing Time Distribution",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 53820,
      "title": "Feature engineering - how to determine good features?",
      "comment_count": 12,
      "views": 0,
      "votes": null
    },
    {
      "id": 53881,
      "title": "(Manageable) Code Structure/Environment for Competitions Code",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 53792,
      "title": "R multicore usage with multidplyr",
      "comment_count": 1,
      "views": 0,
      "votes": 6
    },
    {
      "id": 53832,
      "title": "Complexities of Time series data",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 53754,
      "title": "Are there missing clicks in Train set?",
      "comment_count": 1,
      "views": 0,
      "votes": 6
    },
    {
      "id": 53689,
      "title": "Generative Adversarial Denoising Autoencoder",
      "comment_count": 1,
      "views": 0,
      "votes": 7
    },
    {
      "id": 52805,
      "title": "Anyone use the LightGBM with GPU?",
      "comment_count": 7,
      "views": 0,
      "votes": 5
    },
    {
      "id": 53815,
      "title": "Newbie question: is all this hassle about the IP encoding normal for Kaggle?",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 53250,
      "title": "Is it just me?",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 53567,
      "title": "Data Scientist and a child",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 53846,
      "title": "Data for local experiments",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 53752,
      "title": "Feature Engineering Doubt",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53817,
      "title": "Just curious how this model is to be applied within business system",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53653,
      "title": "Weight of Evidence and Information Value (in Python)",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 53546,
      "title": "Newbie who wants to learn by doing but where to start?",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 53384,
      "title": "proper hour aggregations require test_supplement",
      "comment_count": 12,
      "views": 0,
      "votes": 6
    },
    {
      "id": 53325,
      "title": "How do you test your solution?",
      "comment_count": 28,
      "views": 0,
      "votes": 5
    },
    {
      "id": 52610,
      "title": "This game has a  leak",
      "comment_count": 12,
      "views": 0,
      "votes": 5
    },
    {
      "id": 53574,
      "title": "Extra data hurting LB score",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 53485,
      "title": "How data was prepared for this competition?",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 52802,
      "title": "WHY STOP BLENDS ? Here's what I suggest to Kaggle Admins",
      "comment_count": 16,
      "views": 0,
      "votes": 8
    },
    {
      "id": 53560,
      "title": "XGBoost Python - trouble loading test data ?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53257,
      "title": "Adversarial Validation",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 53458,
      "title": "Which validation set to use?",
      "comment_count": 9,
      "views": 0,
      "votes": 2
    },
    {
      "id": 53521,
      "title": "A question about context.",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 53225,
      "title": "Best single model public lb score?",
      "comment_count": 10,
      "views": 0,
      "votes": null
    },
    {
      "id": 53421,
      "title": "attributed_time?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 52580,
      "title": "Team Merge Issues using Team Name",
      "comment_count": 7,
      "views": 0,
      "votes": 4
    },
    {
      "id": 52868,
      "title": "For NxGTR",
      "comment_count": 11,
      "views": 0,
      "votes": 6
    },
    {
      "id": 53532,
      "title": "XGBoost Factors in R??",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 51245,
      "title": "Do you run code on cloud?",
      "comment_count": 16,
      "views": 0,
      "votes": 10
    },
    {
      "id": 53303,
      "title": "problem with submission format big difference cv/LB",
      "comment_count": 16,
      "views": 0,
      "votes": null
    },
    {
      "id": 53397,
      "title": "More data or More features?",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53249,
      "title": "Best single model?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53274,
      "title": "AutoEncoders, Anyone?",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 52570,
      "title": "Problem forming a team",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 53299,
      "title": "does correlation effect?",
      "comment_count": 7,
      "views": 0,
      "votes": null
    },
    {
      "id": 53188,
      "title": "Should we exhaust all the possibilities of the combination for features ?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 53126,
      "title": "Shuffled split significantly improves the score on LB",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 52561,
      "title": "Not able to merge team",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53111,
      "title": "That feeling ...",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 53283,
      "title": "Getting reproducible results with LGBM",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 52679,
      "title": "Kernels",
      "comment_count": 5,
      "views": 0,
      "votes": 3
    },
    {
      "id": 51348,
      "title": "Encoded Data Questions",
      "comment_count": 6,
      "views": 0,
      "votes": 7
    },
    {
      "id": 52927,
      "title": "From 100k to 100MM rows",
      "comment_count": 4,
      "views": 0,
      "votes": 13
    },
    {
      "id": 51492,
      "title": "Validation",
      "comment_count": 42,
      "views": 0,
      "votes": 28
    },
    {
      "id": 52374,
      "title": "False leak?",
      "comment_count": 24,
      "views": 0,
      "votes": 19
    },
    {
      "id": 53227,
      "title": "About 'click_time'",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 53016,
      "title": "Look at 'Blending' from another perspective",
      "comment_count": 7,
      "views": 0,
      "votes": null
    },
    {
      "id": 53094,
      "title": "Automatically generate statistical features with DL",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 51411,
      "title": "How to manage the amount of data",
      "comment_count": 31,
      "views": 0,
      "votes": 69
    },
    {
      "id": 53136,
      "title": "IP as a predictor or not?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 53130,
      "title": "what does evaluation error 404 mean?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 52999,
      "title": "Data Imbalance",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 53067,
      "title": "What will the private 82% test set look like?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 52985,
      "title": "Why there is no 'server ip address' in Data?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 52752,
      "title": "Duplicate rows in train, test & test-supplement datasets - how come?",
      "comment_count": 5,
      "views": 0,
      "votes": 7
    },
    {
      "id": 52928,
      "title": "SME on Channel",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 52934,
      "title": "R version",
      "comment_count": 15,
      "views": 0,
      "votes": null
    },
    {
      "id": 52854,
      "title": "Why so many Blend Kernel ? ",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 52931,
      "title": "Adding Data source based on kernel version",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 52922,
      "title": "trouble loading the dataset into R.",
      "comment_count": 3,
      "views": 0,
      "votes": 0
    },
    {
      "id": 52890,
      "title": "Using all of the data",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 52833,
      "title": "Public Leaderboard score? ",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 52878,
      "title": "Bad data? Instant Downloads?",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 52933,
      "title": "Pandas on ray",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 52338,
      "title": "ROC function to find your model score",
      "comment_count": 5,
      "views": 0,
      "votes": 6
    },
    {
      "id": 52889,
      "title": "Why is IP address a number?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 52862,
      "title": "What should be the format of the result？",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 52373,
      "title": "Big Test / Small Test - Evaluating Next Steps",
      "comment_count": 12,
      "views": 0,
      "votes": 8
    },
    {
      "id": 52775,
      "title": "Difference between train and test ip",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 52602,
      "title": "How to sort from command line?",
      "comment_count": 12,
      "views": 0,
      "votes": 2
    },
    {
      "id": 52801,
      "title": "Installing XGBoost with GPU support on Windows 10",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 52745,
      "title": "Is it safety to upload submission from Amazon EC2 instances directly?",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 51877,
      "title": "Test data has only half of clicks",
      "comment_count": 19,
      "views": 0,
      "votes": 54
    },
    {
      "id": 52201,
      "title": "Freq counts from Train & Test datasets",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 52786,
      "title": "Old test data survives through blends",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 52637,
      "title": "Not able to add a kaggle member to the team",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 52703,
      "title": "What happened during hours 6,7,8, 11 and 12 in test set?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 52716,
      "title": "How to use scale_pos_weight in LightGBM and XGBoost",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 52721,
      "title": "Why using label encoding on channel, app, device and os ?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 52673,
      "title": "How big is the training dataset?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 52675,
      "title": "Thoughts about the encoded IPs",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 52304,
      "title": "Looking for tips to deal diligently with a large dataset",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 52354,
      "title": "Why different AUC local train/test sets and leaderboard set",
      "comment_count": 8,
      "views": 0,
      "votes": 5
    },
    {
      "id": 52227,
      "title": "LB saturation",
      "comment_count": 11,
      "views": 0,
      "votes": 3
    },
    {
      "id": 52352,
      "title": "Original Test Data vs Halved Test Data",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 52514,
      "title": "Extremely slow scoring",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 52178,
      "title": "Best sample size and how much we can push the kaggle servers?",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 52179,
      "title": "How is it possible that operating systems with the smallest numbers attribute to most of the sales?",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 51162,
      "title": "Ads Clicked & App Downloaded, Then what?",
      "comment_count": 20,
      "views": 0,
      "votes": 18
    },
    {
      "id": 52280,
      "title": "Why I can't invite my partner to join my team?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 52167,
      "title": "Quick tip for subsampling across time when reading the data",
      "comment_count": 1,
      "views": 0,
      "votes": 12
    },
    {
      "id": 51257,
      "title": "DELETED",
      "comment_count": 9,
      "views": 0,
      "votes": null
    },
    {
      "id": 51900,
      "title": "I have a guess that the price of user's phone may play a role in predicting?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 51506,
      "title": "Old test data",
      "comment_count": 6,
      "views": 0,
      "votes": 35
    },
    {
      "id": 51767,
      "title": "Description of Features",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 51504,
      "title": "AWS EC2 advice for running algorithms on whole data",
      "comment_count": 16,
      "views": 0,
      "votes": 2
    },
    {
      "id": 51809,
      "title": "Random selection of all the train set 🎲",
      "comment_count": 4,
      "views": 0,
      "votes": 7
    },
    {
      "id": 52036,
      "title": "Future Feature",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 52265,
      "title": "Label Encoding",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 52053,
      "title": "> 1 hour ? ",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 51914,
      "title": "Let the hunt begin",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 51422,
      "title": "A few model suggestions",
      "comment_count": 3,
      "views": 0,
      "votes": 40
    },
    {
      "id": 51823,
      "title": "Shrinking the data set using Python and SQLite",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 51942,
      "title": "How to deal with hour component of timestamp?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 51514,
      "title": "Extremely slow scoring",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 51373,
      "title": "Prediction type and sample submission file",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 51574,
      "title": "FTLR for C++",
      "comment_count": 0,
      "views": 0,
      "votes": 12
    },
    {
      "id": 52005,
      "title": "Introduce xLearn package for this competion",
      "comment_count": 7,
      "views": 0,
      "votes": -20
    },
    {
      "id": 51670,
      "title": "What Does 'channel' mean?",
      "comment_count": 8,
      "views": 0,
      "votes": 6
    },
    {
      "id": 51432,
      "title": "What exactly is attributed_time?",
      "comment_count": 11,
      "views": 0,
      "votes": 5
    },
    {
      "id": 51818,
      "title": "Display Same ad to Same user after clicking the ad",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 51848,
      "title": "Submission decimal or 0/1",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 51650,
      "title": "why Fraud detection shouldnt be at IP level",
      "comment_count": 6,
      "views": 0,
      "votes": 2
    },
    {
      "id": 51168,
      "title": "Unusual train.csv.zip structure",
      "comment_count": 18,
      "views": 0,
      "votes": 3
    },
    {
      "id": 51715,
      "title": "Why I am not able to load to load the data on kernel?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 51810,
      "title": "IP Addresses - are they real or encoded?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 51236,
      "title": "Data encoding - is it completely random?",
      "comment_count": 8,
      "views": 0,
      "votes": 2
    },
    {
      "id": 51567,
      "title": "Is ip = 1 in train data the same as ip = 1 in test data?",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 51620,
      "title": "Data Talks - Exploring Full dataset",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 51455,
      "title": "Need Advice on handling large volume of data",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 51518,
      "title": "data descriptions",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 51592,
      "title": "Weird Kaggle forum display bug",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 51599,
      "title": "A doubt in Kernals",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 51594,
      "title": "Mork calling Orson",
      "comment_count": 0,
      "views": 0,
      "votes": -1
    },
    {
      "id": 51154,
      "title": "A bitter memory",
      "comment_count": 4,
      "views": 0,
      "votes": 8
    },
    {
      "id": 51441,
      "title": "Julia kernels not supported?",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 51497,
      "title": "Zipped submission file can not be submitted",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 51190,
      "title": "Why there is a gap between training set and test set in click_time?",
      "comment_count": 9,
      "views": 0,
      "votes": 2
    },
    {
      "id": 51400,
      "title": "How to solve ValueError: Length of values does not match length of index?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 51235,
      "title": "Time Features",
      "comment_count": 1,
      "views": 0,
      "votes": 6
    },
    {
      "id": 51369,
      "title": "just 4days/week",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 51279,
      "title": "So much data",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 51150,
      "title": "Data Page",
      "comment_count": 2,
      "views": 0,
      "votes": null
    }
  ],
  "errors": []
}