{
  "id": 175346,
  "title": "Context Specific Submissions thread",
  "url": "/competitions/siim-isic-melanoma-classification/discussion/175346",
  "author_name": "",
  "post_date": "2020-08-18T01:21:21.770265400Z",
  "votes": null,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Can we have a context specific work for this melonma, because it would be really interesting to know about and learn from those submissions to further our knowledge on the different use cases possible . Thank you</p>",
  "messages": [
    {
      "id": "974538",
      "postDate": "08/18/2020 01:21:21",
      "content": "<p>Can we have a context specific work for this melonma, because it would be really interesting to know about and learn from those submissions to further our knowledge on the different use cases possible . Thank you</p>",
      "rawMarkdown": "Can we have a context specific work for this melonma, because it would be really interesting to know about and learn from those submissions to further our knowledge on the different use cases possible . Thank you",
      "votes": null
    },
    {
      "id": "976613",
      "postDate": "08/19/2020 01:29:53",
      "content": "<p>Iam getting upvotes for this thread! But it is not reflected here! Any issues with Kaggle system</p>",
      "rawMarkdown": "Iam getting upvotes for this thread! But it is not reflected here! Any issues with Kaggle system",
      "votes": null
    },
    {
      "id": "976629",
      "postDate": "08/19/2020 02:00:54",
      "content": "<p>I got a boost on CV using the feature \"number of images per patient\", i.e. <code>train['ct'] = train.groupby('patient_id').patient_id.transform('count')</code>. Besides that, I didn't increase CV using any other context specific features. </p>\n<p>I tried many things including ugly duckling and trying to assesss evolution based on <code>patient_id</code> and <code>age</code> but didn't get either to work.</p>",
      "rawMarkdown": "I got a boost on CV using the feature \"number of images per patient\", i.e. `train['ct'] = train.groupby('patient_id').patient_id.transform('count')`. Besides that, I didn't increase CV using any other context specific features. \n\nI tried many things including ugly duckling and trying to assesss evolution based on `patient_id` and `age` but didn't get either to work.",
      "votes": null
    },
    {
      "id": "976659",
      "postDate": "08/19/2020 02:45:09",
      "content": "<p>Thanks Chris Deotte, Nice to Hear</p>",
      "rawMarkdown": "Thanks Chris Deotte, Nice to Hear",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 976613,
      "author_name": "loveall",
      "author_url": "",
      "post_date": "08/19/2020 01:29:53",
      "content": "<p>Iam getting upvotes for this thread! But it is not reflected here! Any issues with Kaggle system</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 976629,
      "author_name": "cdeotte",
      "author_url": "",
      "post_date": "08/19/2020 02:00:54",
      "content": "<p>I got a boost on CV using the feature \"number of images per patient\", i.e. <code>train['ct'] = train.groupby('patient_id').patient_id.transform('count')</code>. Besides that, I didn't increase CV using any other context specific features. </p>\n<p>I tried many things including ugly duckling and trying to assesss evolution based on <code>patient_id</code> and <code>age</code> but didn't get either to work.</p>",
      "votes": null,
      "replies": [
        {
          "id": 976659,
          "author_name": "loveall",
          "author_url": "",
          "post_date": "08/19/2020 02:45:09",
          "content": "<p>Thanks Chris Deotte, Nice to Hear</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "974538": "Can we have a context specific work for this melonma, because it would be really interesting to know about and learn from those submissions to further our knowledge on the different use cases possible . Thank you",
    "976613": "Iam getting upvotes for this thread! But it is not reflected here! Any issues with Kaggle system",
    "976629": "I got a boost on CV using the feature \"number of images per patient\", i.e. `train['ct'] = train.groupby('patient_id').patient_id.transform('count')`. Besides that, I didn't increase CV using any other context specific features. \n\nI tried many things including ugly duckling and trying to assesss evolution based on `patient_id` and `age` but didn't get either to work.",
    "976659": "Thanks Chris Deotte, Nice to Hear"
  },
  "source": "meta"
}