{
  "id": 153276,
  "title": "Has anyone used gleason_score for classification? I said my thoughts",
  "url": "/competitions/prostate-cancer-grade-assessment/discussion/153276",
  "author_name": "zhangeng",
  "post_date": "2020-05-24T01:54:21.505000",
  "votes": 3,
  "comment_count": 1,
  "views": 0,
  "content": "<p>1.gleason_score has 4 single categories(0,3,4,5) and 6 combinations(0+0,3+3,3+4,4+3,4+4,(4+5,5+5)).</p>\n\n<p>2.We use the gleason_score  to first divide into 4 single categories, and then select the 2 most probable categories from the 4 single categories to combine, and then convert into 6 combined categories (isup_grade) after scoring.</p>\n\n<p>3.Just provide ideas, have not tried</p>",
  "messages": [
    {
      "id": 858961,
      "postDate": "2020-05-24T01:54:21.507Z",
      "content": "<p>1.gleason_score has 4 single categories(0,3,4,5) and 6 combinations(0+0,3+3,3+4,4+3,4+4,(4+5,5+5)).</p>\n\n<p>2.We use the gleason_score  to first divide into 4 single categories, and then select the 2 most probable categories from the 4 single categories to combine, and then convert into 6 combined categories (isup_grade) after scoring.</p>\n\n<p>3.Just provide ideas, have not tried</p>",
      "rawMarkdown": "1.gleason_score has 4 single categories(0,3,4,5) and 6 combinations(0+0,3+3,3+4,4+3,4+4,(4+5,5+5)).\n\n2.We use the gleason_score  to first divide into 4 single categories, and then select the 2 most probable categories from the 4 single categories to combine, and then convert into 6 combined categories (isup_grade) after scoring.\n\n3.Just provide ideas, have not tried",
      "votes": 3
    },
    {
      "id": 859537,
      "postDate": "2020-05-24T14:27:36.783Z",
      "content": "<p>Its a grading system for checking the severity of prostate cancer. The score lies between 1 to 5, 1 being the most healthier and 5 being the most cancerous.\nYou could use it in your classification problem but if you don't already have it in the data, you could use clustering algos to label them to different classes from 1 to 5.</p>",
      "rawMarkdown": "Its a grading system for checking the severity of prostate cancer. The score lies between 1 to 5, 1 being the most healthier and 5 being the most cancerous.\nYou could use it in your classification problem but if you don't already have it in the data, you could use clustering algos to label them to different classes from 1 to 5."
    }
  ],
  "comments": [
    {
      "id": 859537,
      "author_name": "Pratik Asarkar",
      "author_url": "",
      "post_date": "2020-05-24T14:27:36.783000",
      "content": "<p>Its a grading system for checking the severity of prostate cancer. The score lies between 1 to 5, 1 being the most healthier and 5 being the most cancerous.\nYou could use it in your classification problem but if you don't already have it in the data, you could use clustering algos to label them to different classes from 1 to 5.</p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "858961": "1.gleason_score has 4 single categories(0,3,4,5) and 6 combinations(0+0,3+3,3+4,4+3,4+4,(4+5,5+5)).\n\n2.We use the gleason_score  to first divide into 4 single categories, and then select the 2 most probable categories from the 4 single categories to combine, and then convert into 6 combined categories (isup_grade) after scoring.\n\n3.Just provide ideas, have not tried",
    "859537": "Its a grading system for checking the severity of prostate cancer. The score lies between 1 to 5, 1 being the most healthier and 5 being the most cancerous.\nYou could use it in your classification problem but if you don't already have it in the data, you could use clustering algos to label them to different classes from 1 to 5."
  }
}