{
  "id": 222324,
  "title": "Is the competition setup realistic for applications?",
  "url": "/competitions/indoor-location-navigation/discussion/222324",
  "author_name": "",
  "post_date": "2021-02-26T10:56:58.019917800Z",
  "votes": 11,
  "comment_count": 1,
  "views": 0,
  "content": "<p>The data provided in competitions contains walking paths with their timestamps. So, I am guessing that the best solution will be some kind of <strong>bidirectional</strong> model that takes into account both past and future data. However, is this realistic for a real application? In reality, this bidirectionality is not genuine and we would be forced to use just past data. <br>\nI think it is a pity that at the end of the competition we found out that the best strategies can not be transitioned into real applications.</p>",
  "messages": [
    {
      "id": "1218995",
      "postDate": "02/26/2021 10:56:58",
      "content": "<p>The data provided in competitions contains walking paths with their timestamps. So, I am guessing that the best solution will be some kind of <strong>bidirectional</strong> model that takes into account both past and future data. However, is this realistic for a real application? In reality, this bidirectionality is not genuine and we would be forced to use just past data. <br>\nI think it is a pity that at the end of the competition we found out that the best strategies can not be transitioned into real applications.</p>",
      "rawMarkdown": "The data provided in competitions contains walking paths with their timestamps. So, I am guessing that the best solution will be some kind of **bidirectional** model that takes into account both past and future data. However, is this realistic for a real application? In reality, this bidirectionality is not genuine and we would be forced to use just past data. \nI think it is a pity that at the end of the competition we found out that the best strategies can not be transitioned into real applications.",
      "votes": null
    },
    {
      "id": "1231574",
      "postDate": "03/09/2021 04:47:55",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/joelqv\" target=\"_blank\">@joelqv</a>, I agree with you that model using future data might have difficulty in translating to real-world application. But  I guess decent results can be obtained using just the past data alone.</p>",
      "rawMarkdown": "Hi @joelqv, I agree with you that model using future data might have difficulty in translating to real-world application. But  I guess decent results can be obtained using just the past data alone.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1231574,
      "author_name": "suryajrrafl",
      "author_url": "",
      "post_date": "03/09/2021 04:47:55",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/joelqv\" target=\"_blank\">@joelqv</a>, I agree with you that model using future data might have difficulty in translating to real-world application. But  I guess decent results can be obtained using just the past data alone.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1218995": "The data provided in competitions contains walking paths with their timestamps. So, I am guessing that the best solution will be some kind of **bidirectional** model that takes into account both past and future data. However, is this realistic for a real application? In reality, this bidirectionality is not genuine and we would be forced to use just past data. \nI think it is a pity that at the end of the competition we found out that the best strategies can not be transitioned into real applications.",
    "1231574": "Hi @joelqv, I agree with you that model using future data might have difficulty in translating to real-world application. But  I guess decent results can be obtained using just the past data alone."
  },
  "source": "meta"
}