{
  "id": 106150,
  "title": "Performing Error Analysis",
  "url": "/competitions/recursion-cellular-image-classification/discussion/106150",
  "author_name": "wjsheng",
  "post_date": "2019-08-28T12:49:38.926000",
  "votes": 2,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi guys,</p>\n\n<p>I wonder whether is it common to perform error analysis on image classification task such as this competition. For instance, looking at images in the validation set that has the highest errors and do something to improve our model.</p>\n\n<p>Any tip is highly appreciated. Thanks!</p>",
  "messages": [
    {
      "id": 610110,
      "postDate": "2019-08-28T12:49:38.927Z",
      "content": "<p>Hi guys,</p>\n\n<p>I wonder whether is it common to perform error analysis on image classification task such as this competition. For instance, looking at images in the validation set that has the highest errors and do something to improve our model.</p>\n\n<p>Any tip is highly appreciated. Thanks!</p>",
      "rawMarkdown": "Hi guys,\n\nI wonder whether is it common to perform error analysis on image classification task such as this competition. For instance, looking at images in the validation set that has the highest errors and do something to improve our model.\n\nAny tip is highly appreciated. Thanks!",
      "votes": 2
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "610110": "Hi guys,\n\nI wonder whether is it common to perform error analysis on image classification task such as this competition. For instance, looking at images in the validation set that has the highest errors and do something to improve our model.\n\nAny tip is highly appreciated. Thanks!"
  }
}