{
  "id": 110360,
  "title": "Is hand labelling permitted?",
  "url": "/competitions/rsna-intracranial-hemorrhage-detection/discussion/110360",
  "author_name": "Shubhang",
  "post_date": "2019-09-27T04:16:31.337000",
  "votes": 0,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I'd like to try a unusual approach, but I'd need to hand label some of the data, is that permitted?</p>",
  "messages": [
    {
      "id": 635049,
      "postDate": "2019-09-27T05:21:57.437Z",
      "content": "<p>Looks like you're free to annotate and label the training data, but not the test data.</p>\n\n<p>From the rules section: </p>\n\n<blockquote>\n  <p>Entrants may re-annotate images in the training set, but may not hand-label predictions, including having human observers rate and evaluate the test data set.</p>\n</blockquote>",
      "rawMarkdown": "Looks like you're free to annotate and label the training data, but not the test data.\n\nFrom the rules section: \n&gt; Entrants may re-annotate images in the training set, but may not hand-label predictions, including having human observers rate and evaluate the test data set.",
      "votes": 3
    },
    {
      "id": 635210,
      "postDate": "2019-09-27T08:57:55.667Z",
      "content": "<p>That would a bit too much advantage if one re-annotates the data using expertise of a professional radiologist, I believe :)</p>",
      "rawMarkdown": "That would a bit too much advantage if one re-annotates the data using expertise of a professional radiologist, I believe :)",
      "votes": 1
    },
    {
      "id": 635017,
      "postDate": "2019-09-27T04:16:31.337Z",
      "content": "<p>I'd like to try a unusual approach, but I'd need to hand label some of the data, is that permitted?</p>",
      "rawMarkdown": "I'd like to try a unusual approach, but I'd need to hand label some of the data, is that permitted?"
    }
  ],
  "comments": [
    {
      "id": 635049,
      "author_name": "Ryan Epp",
      "author_url": "",
      "post_date": "2019-09-27T05:21:57.437000",
      "content": "<p>Looks like you're free to annotate and label the training data, but not the test data.</p>\n\n<p>From the rules section: </p>\n\n<blockquote>\n  <p>Entrants may re-annotate images in the training set, but may not hand-label predictions, including having human observers rate and evaluate the test data set.</p>\n</blockquote>",
      "votes": 3,
      "replies": []
    },
    {
      "id": 635210,
      "author_name": "Ilia Zaitsev",
      "author_url": "",
      "post_date": "2019-09-27T08:57:55.667000",
      "content": "<p>That would a bit too much advantage if one re-annotates the data using expertise of a professional radiologist, I believe :)</p>",
      "votes": 1,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "635049": "Looks like you're free to annotate and label the training data, but not the test data.\n\nFrom the rules section: \n&gt; Entrants may re-annotate images in the training set, but may not hand-label predictions, including having human observers rate and evaluate the test data set.",
    "635210": "That would a bit too much advantage if one re-annotates the data using expertise of a professional radiologist, I believe :)",
    "635017": "I'd like to try a unusual approach, but I'd need to hand label some of the data, is that permitted?"
  }
}