{
  "id": 21279,
  "title": "Training Data - Filtering",
  "url": "/competitions/expedia-hotel-recommendations/discussion/21279",
  "author_name": "",
  "post_date": "2016-05-28T09:32:03.397Z",
  "votes": 1,
  "comment_count": 1,
  "views": 543,
  "content": "<p>Hi Everyone,</p>\n\n<p>From train data,I'm considering rows where Booking equals '1' i.e., is_booking = 1\nfor developing model. Is it correct way ? Can someone confirm me.</p>\n\n<p>Regards,\nSiva </p>",
  "messages": [
    {
      "id": "121642",
      "postDate": "05/28/2016 09:32:03",
      "content": "<p>Hi Everyone,</p>\n\n<p>From train data,I'm considering rows where Booking equals '1' i.e., is_booking = 1\nfor developing model. Is it correct way ? Can someone confirm me.</p>\n\n<p>Regards,\nSiva </p>",
      "rawMarkdown": "Hi Everyone,\r\n\r\nFrom train data,I'm considering rows where Booking equals '1' i.e., is_booking = 1\r\nfor developing model. Is it correct way ? Can someone confirm me.\r\n\r\nRegards,\r\nSiva",
      "votes": null
    },
    {
      "id": "122004",
      "postDate": "05/31/2016 15:45:07",
      "content": "<p>Hi Siva,\nI think that it depends to your algorithm.</p>\n\n<p>In my undersatnding, rows where Booking equals '1' i.e., is_booking = 1 are the very usefull for your prediction as they give the users' final choices. But considering &quot;click-only&quot; rows may be interesting. The fact a user clicks on an hotel_cluster signals a certain interest she has on it.</p>\n\n<p>--\nModou</p>",
      "rawMarkdown": "Hi Siva,\r\nI think that it depends to your algorithm.\r\n\r\nIn my undersatnding, rows where Booking equals '1' i.e., is_booking = 1 are the very usefull for your prediction as they give the users' final choices. But considering \"click-only\" rows may be interesting. The fact a user clicks on an hotel_cluster signals a certain interest she has on it.\r\n\r\n--\r\nModou",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 122004,
      "author_name": "mogueye",
      "author_url": "",
      "post_date": "05/31/2016 15:45:07",
      "content": "<p>Hi Siva,\nI think that it depends to your algorithm.</p>\n\n<p>In my undersatnding, rows where Booking equals '1' i.e., is_booking = 1 are the very usefull for your prediction as they give the users' final choices. But considering &quot;click-only&quot; rows may be interesting. The fact a user clicks on an hotel_cluster signals a certain interest she has on it.</p>\n\n<p>--\nModou</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "121642": "Hi Everyone,\r\n\r\nFrom train data,I'm considering rows where Booking equals '1' i.e., is_booking = 1\r\nfor developing model. Is it correct way ? Can someone confirm me.\r\n\r\nRegards,\r\nSiva",
    "122004": "Hi Siva,\r\nI think that it depends to your algorithm.\r\n\r\nIn my undersatnding, rows where Booking equals '1' i.e., is_booking = 1 are the very usefull for your prediction as they give the users' final choices. But considering \"click-only\" rows may be interesting. The fact a user clicks on an hotel_cluster signals a certain interest she has on it.\r\n\r\n--\r\nModou"
  },
  "source": "meta"
}