{
  "id": 176911,
  "title": "Thoughts on shakeup?",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/176911",
  "author_name": "",
  "post_date": "2020-08-24T05:18:15.281235800Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Someone mentioned about the skew data, would be there any shakeup?</p>",
  "messages": [
    {
      "id": "983181",
      "postDate": "08/24/2020 05:18:15",
      "content": "<p>Someone mentioned about the skew data, would be there any shakeup?</p>",
      "rawMarkdown": "Someone mentioned about the skew data, would be there any shakeup?",
      "votes": null
    },
    {
      "id": "983215",
      "postDate": "08/24/2020 05:38:41",
      "content": "<p>Definitely a shakeup in my opinion.</p>",
      "rawMarkdown": "Definitely a shakeup in my opinion.",
      "votes": null
    },
    {
      "id": "983397",
      "postDate": "08/24/2020 09:18:42",
      "content": "<p>There is gonna be a huge shakeup:</p>\n<ol>\n<li>15/85 public/private split</li>\n<li>There are outliers in train data</li>\n<li>Two characteristics at once - FVC and confidence - easy to overfit model on any of this 'parameters'</li>\n<li>A lot of top public notebooks with -6.8 public LB</li>\n</ol>",
      "rawMarkdown": "There is gonna be a huge shakeup:\n1. 15/85 public/private split\n2. There are outliers in train data\n3. Two characteristics at once - FVC and confidence - easy to overfit model on any of this 'parameters'\n4. A lot of top public notebooks with -6.8 public LB",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 983215,
      "author_name": "matthewmasters",
      "author_url": "",
      "post_date": "08/24/2020 05:38:41",
      "content": "<p>Definitely a shakeup in my opinion.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 983397,
      "author_name": "eliasgreen",
      "author_url": "",
      "post_date": "08/24/2020 09:18:42",
      "content": "<p>There is gonna be a huge shakeup:</p>\n<ol>\n<li>15/85 public/private split</li>\n<li>There are outliers in train data</li>\n<li>Two characteristics at once - FVC and confidence - easy to overfit model on any of this 'parameters'</li>\n<li>A lot of top public notebooks with -6.8 public LB</li>\n</ol>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "983181": "Someone mentioned about the skew data, would be there any shakeup?",
    "983215": "Definitely a shakeup in my opinion.",
    "983397": "There is gonna be a huge shakeup:\n1. 15/85 public/private split\n2. There are outliers in train data\n3. Two characteristics at once - FVC and confidence - easy to overfit model on any of this 'parameters'\n4. A lot of top public notebooks with -6.8 public LB"
  },
  "source": "meta"
}