{
  "id": 59865,
  "title": "Good luck )) I hope in future we'll see many interest competitions, where different areas of Machine learning meet. It is very interest to work both with tabular data, natural language processing and computer vision problem at the same time",
  "url": "/competitions/avito-demand-prediction/discussion/59865",
  "author_name": "",
  "post_date": "2018-06-27T21:53:19.197870200Z",
  "votes": 23,
  "comment_count": 1,
  "views": 0,
  "content": "",
  "messages": [
    {
      "id": "349154",
      "postDate": "06/27/2018 21:53:19",
      "content": "",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "350660",
      "postDate": "06/30/2018 07:34:00",
      "content": "<p>I think I have found another thing that I am not too keen on with Kernel competitions. As I've experimented with various seeds, I've noticed that that the overall performance of the algorithm can vary substantially. One lucky seed can make a script appear to perform much better than it really should. This is usually confined to the third decimal though. In the real world this may not be that much of an issue, but in Kaggle competitions that can be a real deal breaker <a href=\"https://www.essaytime.co.uk/\">Essay Help Online</a></p>",
      "rawMarkdown": "I think I have found another thing that I am not too keen on with Kernel competitions. As I've experimented with various seeds, I've noticed that that the overall performance of the algorithm can vary substantially. One lucky seed can make a script appear to perform much better than it really should. This is usually confined to the third decimal though. In the real world this may not be that much of an issue, but in Kaggle competitions that can be a real deal breaker [Essay Help Online][1]\n\n\n  [1]: https://www.essaytime.co.uk/",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 350660,
      "author_name": "elianahermione",
      "author_url": "",
      "post_date": "06/30/2018 07:34:00",
      "content": "<p>I think I have found another thing that I am not too keen on with Kernel competitions. As I've experimented with various seeds, I've noticed that that the overall performance of the algorithm can vary substantially. One lucky seed can make a script appear to perform much better than it really should. This is usually confined to the third decimal though. In the real world this may not be that much of an issue, but in Kaggle competitions that can be a real deal breaker <a href=\"https://www.essaytime.co.uk/\">Essay Help Online</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "349154": "",
    "350660": "I think I have found another thing that I am not too keen on with Kernel competitions. As I've experimented with various seeds, I've noticed that that the overall performance of the algorithm can vary substantially. One lucky seed can make a script appear to perform much better than it really should. This is usually confined to the third decimal though. In the real world this may not be that much of an issue, but in Kaggle competitions that can be a real deal breaker [Essay Help Online][1]\n\n\n  [1]: https://www.essaytime.co.uk/"
  },
  "source": "meta"
}