{
  "id": 90123,
  "title": "PublicLB are evaluated based on only  13% of the test data.",
  "url": "/competitions/LANL-Earthquake-Prediction/discussion/90123",
  "author_name": "",
  "post_date": "2019-04-20T16:44:29.512741600Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Public LB are evaluated based on only 13% of the test data. I think this is a very small percentage. In this case, what are you aware of? And, what kind of technique do you think is effective to avoid overfitting in such a case?</p>",
  "messages": [
    {
      "id": "520300",
      "postDate": "04/20/2019 16:44:29",
      "content": "<p>Public LB are evaluated based on only 13% of the test data. I think this is a very small percentage. In this case, what are you aware of? And, what kind of technique do you think is effective to avoid overfitting in such a case?</p>",
      "rawMarkdown": "Public LB are evaluated based on only 13% of the test data. I think this is a very small percentage. In this case, what are you aware of? And, what kind of technique do you think is effective to avoid overfitting in such a case?",
      "votes": null
    },
    {
      "id": "520333",
      "postDate": "04/20/2019 18:14:31",
      "content": "<p>The good old technique named \"trust your CV\" </p>",
      "rawMarkdown": "The good old technique named \"trust your CV\"",
      "votes": null
    },
    {
      "id": "520337",
      "postDate": "04/20/2019 18:29:28",
      "content": "<p>Thank you!  I have heard about that old technique \"trust your CV\" :)\nMay the force be with you.</p>",
      "rawMarkdown": "Thank you!  I have heard about that old technique \"trust your CV\" :)\nMay the force be with you.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 520333,
      "author_name": "stecasasso",
      "author_url": "",
      "post_date": "04/20/2019 18:14:31",
      "content": "<p>The good old technique named \"trust your CV\" </p>",
      "votes": null,
      "replies": [
        {
          "id": 520337,
          "author_name": "dhaqui",
          "author_url": "",
          "post_date": "04/20/2019 18:29:28",
          "content": "<p>Thank you!  I have heard about that old technique \"trust your CV\" :)\nMay the force be with you.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "520300": "Public LB are evaluated based on only 13% of the test data. I think this is a very small percentage. In this case, what are you aware of? And, what kind of technique do you think is effective to avoid overfitting in such a case?",
    "520333": "The good old technique named \"trust your CV\"",
    "520337": "Thank you!  I have heard about that old technique \"trust your CV\" :)\nMay the force be with you."
  },
  "source": "meta"
}