{
  "id": 66163,
  "title": "Validation Set",
  "url": "/competitions/ga-customer-revenue-prediction/discussion/66163",
  "author_name": "",
  "post_date": "2018-09-18T21:13:19.239293600Z",
  "votes": 2,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hello Kagglers,</p>\n\n<p>I would like to hear your thoughts about the validation set. At the moment I am in dilemma between the following options:\n1) Divide based on full visotor ID\n2) Divide based on dates</p>\n\n<p>As our target is to make predictions for the next year based on last years data I will probably go with the first option.</p>\n\n<p>Happy to hear your thoughts!!\nGood Luck!! </p>",
  "messages": [
    {
      "id": "389519",
      "postDate": "09/18/2018 21:13:19",
      "content": "<p>Hello Kagglers,</p>\n\n<p>I would like to hear your thoughts about the validation set. At the moment I am in dilemma between the following options:\n1) Divide based on full visotor ID\n2) Divide based on dates</p>\n\n<p>As our target is to make predictions for the next year based on last years data I will probably go with the first option.</p>\n\n<p>Happy to hear your thoughts!!\nGood Luck!! </p>",
      "rawMarkdown": "Hello Kagglers,\n\nI would like to hear your thoughts about the validation set. At the moment I am in dilemma between the following options:\n1) Divide based on full visotor ID\n2) Divide based on dates\n\nAs our target is to make predictions for the next year based on last years data I will probably go with the first option.\n\nHappy to hear your thoughts!!\nGood Luck!!",
      "votes": null
    },
    {
      "id": "399728",
      "postDate": "10/06/2018 15:41:42",
      "content": "<p>I would prefer dates - same because of the train-test is splitted.</p>",
      "rawMarkdown": "I would prefer dates - same because of the train-test is splitted.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 399728,
      "author_name": "hemantrupani",
      "author_url": "",
      "post_date": "10/06/2018 15:41:42",
      "content": "<p>I would prefer dates - same because of the train-test is splitted.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "389519": "Hello Kagglers,\n\nI would like to hear your thoughts about the validation set. At the moment I am in dilemma between the following options:\n1) Divide based on full visotor ID\n2) Divide based on dates\n\nAs our target is to make predictions for the next year based on last years data I will probably go with the first option.\n\nHappy to hear your thoughts!!\nGood Luck!!",
    "399728": "I would prefer dates - same because of the train-test is splitted."
  },
  "source": "meta"
}