{
  "id": 193997,
  "title": "Are there 780K - 400K = 380K new users in the hidden data?",
  "url": "/competitions/riiid-test-answer-prediction/discussion/193997",
  "author_name": "",
  "post_date": "2020-10-30T05:47:29.924054500Z",
  "votes": 3,
  "comment_count": 4,
  "views": 0,
  "content": "<p>The riiid! website says:<br>\n\"Build an algorithm to track knowledge states of 780K+ students and win 100K prize\".</p>\n<p>There are almost 400K students in the Training dataset. Are the other 380K students new in the hidden dataset?</p>\n<p>There are 2.5 million interactions in the hidden dataset, so this averages, at most, 6 interactions for each new student, probably a lot less.</p>\n<p>What do you think about this arithmetic??</p>",
  "messages": [
    {
      "id": "1064435",
      "postDate": "10/30/2020 05:47:29",
      "content": "<p>The riiid! website says:<br>\n\"Build an algorithm to track knowledge states of 780K+ students and win 100K prize\".</p>\n<p>There are almost 400K students in the Training dataset. Are the other 380K students new in the hidden dataset?</p>\n<p>There are 2.5 million interactions in the hidden dataset, so this averages, at most, 6 interactions for each new student, probably a lot less.</p>\n<p>What do you think about this arithmetic??</p>",
      "rawMarkdown": "The riiid! website says:\n\"Build an algorithm to track knowledge states of 780K+ students and win 100K prize\".\n\nThere are almost 400K students in the Training dataset. Are the other 380K students new in the hidden dataset?\n\nThere are 2.5 million interactions in the hidden dataset, so this averages, at most, 6 interactions for each new student, probably a lot less.\n\nWhat do you think about this arithmetic??",
      "votes": null
    },
    {
      "id": "1064440",
      "postDate": "10/30/2020 05:55:12",
      "content": "<p>There are <strong>131M</strong> interactions in the original dataset while only <strong>101M</strong> in the Kaggle dataset so a direct comparison is not appropriate.</p>\n<p>It's possible some users (maybe deleted users or those with very high/less interactions) were removed before preparing the dataset.</p>",
      "rawMarkdown": "There are **131M** interactions in the original dataset while only **101M** in the Kaggle dataset so a direct comparison is not appropriate.\n\nIt's possible some users (maybe deleted users or those with very high/less interactions) were removed before preparing the dataset.",
      "votes": null
    },
    {
      "id": "1065026",
      "postDate": "10/30/2020 19:04:37",
      "content": "<blockquote>\n  <p>What do you think about this arithmetic? </p>\n</blockquote>\n<p>I doubt anyone can win 100k. Since this amount is divided between the top 5 places</p>",
      "rawMarkdown": "> What do you think about this arithmetic? \n\nI doubt anyone can win 100k. Since this amount is divided between the top 5 places",
      "votes": null
    },
    {
      "id": "1065093",
      "postDate": "10/30/2020 21:00:00",
      "content": "<p>The headline of this competition is</p>\n<p>Riiid! Answer Correctness Prediction<br>\nTrack knowledge states of <strong>1M+</strong> students in the wild</p>",
      "rawMarkdown": "The headline of this competition is\n\nRiiid! Answer Correctness Prediction\nTrack knowledge states of **1M+** students in the wild",
      "votes": null
    },
    {
      "id": "1065134",
      "postDate": "10/30/2020 22:07:38",
      "content": "<p>Now that could make sense:<br>\n131M interactions -&gt; 1M+ students<br>\n101M interactions -&gt; 780K students</p>",
      "rawMarkdown": "Now that could make sense:\n131M interactions -> 1M+ students\n101M interactions -> 780K students",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1064440,
      "author_name": "rohanrao",
      "author_url": "",
      "post_date": "10/30/2020 05:55:12",
      "content": "<p>There are <strong>131M</strong> interactions in the original dataset while only <strong>101M</strong> in the Kaggle dataset so a direct comparison is not appropriate.</p>\n<p>It's possible some users (maybe deleted users or those with very high/less interactions) were removed before preparing the dataset.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1065026,
      "author_name": "sapr3s",
      "author_url": "",
      "post_date": "10/30/2020 19:04:37",
      "content": "<blockquote>\n  <p>What do you think about this arithmetic? </p>\n</blockquote>\n<p>I doubt anyone can win 100k. Since this amount is divided between the top 5 places</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1065093,
      "author_name": "ilu000",
      "author_url": "",
      "post_date": "10/30/2020 21:00:00",
      "content": "<p>The headline of this competition is</p>\n<p>Riiid! Answer Correctness Prediction<br>\nTrack knowledge states of <strong>1M+</strong> students in the wild</p>",
      "votes": null,
      "replies": [
        {
          "id": 1065134,
          "author_name": "mikel1",
          "author_url": "",
          "post_date": "10/30/2020 22:07:38",
          "content": "<p>Now that could make sense:<br>\n131M interactions -&gt; 1M+ students<br>\n101M interactions -&gt; 780K students</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1064435": "The riiid! website says:\n\"Build an algorithm to track knowledge states of 780K+ students and win 100K prize\".\n\nThere are almost 400K students in the Training dataset. Are the other 380K students new in the hidden dataset?\n\nThere are 2.5 million interactions in the hidden dataset, so this averages, at most, 6 interactions for each new student, probably a lot less.\n\nWhat do you think about this arithmetic??",
    "1064440": "There are **131M** interactions in the original dataset while only **101M** in the Kaggle dataset so a direct comparison is not appropriate.\n\nIt's possible some users (maybe deleted users or those with very high/less interactions) were removed before preparing the dataset.",
    "1065026": "> What do you think about this arithmetic? \n\nI doubt anyone can win 100k. Since this amount is divided between the top 5 places",
    "1065093": "The headline of this competition is\n\nRiiid! Answer Correctness Prediction\nTrack knowledge states of **1M+** students in the wild",
    "1065134": "Now that could make sense:\n131M interactions -> 1M+ students\n101M interactions -> 780K students"
  },
  "source": "meta"
}