{
  "id": 301998,
  "title": "Ensemble learning",
  "url": "/competitions/g-research-crypto-forecasting/discussion/301998",
  "author_name": "",
  "post_date": "2022-01-20T11:31:06.287887800Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>If I use one complicated model and one easy model in ensemble learning, will the result be better than a single complicate model?</p>",
  "messages": [
    {
      "id": "1657680",
      "postDate": "01/20/2022 11:31:06",
      "content": "<p>If I use one complicated model and one easy model in ensemble learning, will the result be better than a single complicate model?</p>",
      "rawMarkdown": "If I use one complicated model and one easy model in ensemble learning, will the result be better than a single complicate model?",
      "votes": null
    },
    {
      "id": "1658284",
      "postDate": "01/20/2022 22:25:24",
      "content": "<p>You must try different ratios of each on multiple test sets to find out how much you can improve. Sometimes yes and sometimes no for me. </p>",
      "rawMarkdown": "You must try different ratios of each on multiple test sets to find out how much you can improve. Sometimes yes and sometimes no for me.",
      "votes": null
    },
    {
      "id": "1659337",
      "postDate": "01/21/2022 18:02:27",
      "content": "<p>I will say it is likely to be overfitting. The Crypto market changes rapidly, and the best ensemble you make may end up with bad predictions in the future.</p>",
      "rawMarkdown": "I will say it is likely to be overfitting. The Crypto market changes rapidly, and the best ensemble you make may end up with bad predictions in the future.",
      "votes": null
    },
    {
      "id": "1665123",
      "postDate": "01/26/2022 15:37:46",
      "content": "<p>Thank you for your advice!</p>",
      "rawMarkdown": "Thank you for your advice!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1658284,
      "author_name": "chrisrichardmiles",
      "author_url": "",
      "post_date": "01/20/2022 22:25:24",
      "content": "<p>You must try different ratios of each on multiple test sets to find out how much you can improve. Sometimes yes and sometimes no for me. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1659337,
      "author_name": "sihead",
      "author_url": "",
      "post_date": "01/21/2022 18:02:27",
      "content": "<p>I will say it is likely to be overfitting. The Crypto market changes rapidly, and the best ensemble you make may end up with bad predictions in the future.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1665123,
      "author_name": "huaiguaiguai",
      "author_url": "",
      "post_date": "01/26/2022 15:37:46",
      "content": "<p>Thank you for your advice!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1657680": "If I use one complicated model and one easy model in ensemble learning, will the result be better than a single complicate model?",
    "1658284": "You must try different ratios of each on multiple test sets to find out how much you can improve. Sometimes yes and sometimes no for me.",
    "1659337": "I will say it is likely to be overfitting. The Crypto market changes rapidly, and the best ensemble you make may end up with bad predictions in the future.",
    "1665123": "Thank you for your advice!"
  },
  "source": "meta"
}