{
  "id": 72433,
  "title": "Which metrics can be used for this task?",
  "url": "/competitions/histopathologic-cancer-detection/discussion/72433",
  "author_name": "",
  "post_date": "2018-11-23T07:44:23.423160500Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "",
  "messages": [
    {
      "id": "426407",
      "postDate": "11/23/2018 07:44:23",
      "content": "",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "426719",
      "postDate": "11/23/2018 17:46:29",
      "content": "<p>Area under the ROC curve between the predicted probability and the observed target.</p>\n\n<p>This is mentioned on the <a href=\"https://www.kaggle.com/c/histopathologic-cancer-detection#evaluation\">Overview Page </a> under <strong>Evaluation</strong> Tab</p>",
      "rawMarkdown": "Area under the ROC curve between the predicted probability and the observed target.\n\nThis is mentioned on the [Overview Page ](https://www.kaggle.com/c/histopathologic-cancer-detection#evaluation) under __Evaluation__ Tab",
      "votes": null
    },
    {
      "id": "430841",
      "postDate": "12/01/2018 03:37:10",
      "content": "<p>Thanks. I have used it in my kernel, but it's wrong when i train my model, i got auc with nan in the last 1000 mini-batch of every epoch, but everything else is ok, like loss, acc. </p>",
      "rawMarkdown": "Thanks. I have used it in my kernel, but it's wrong when i train my model, i got auc with nan in the last 1000 mini-batch of every epoch, but everything else is ok, like loss, acc.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 426719,
      "author_name": "pratik2901",
      "author_url": "",
      "post_date": "11/23/2018 17:46:29",
      "content": "<p>Area under the ROC curve between the predicted probability and the observed target.</p>\n\n<p>This is mentioned on the <a href=\"https://www.kaggle.com/c/histopathologic-cancer-detection#evaluation\">Overview Page </a> under <strong>Evaluation</strong> Tab</p>",
      "votes": null,
      "replies": [
        {
          "id": 430841,
          "author_name": "zhlee2017",
          "author_url": "",
          "post_date": "12/01/2018 03:37:10",
          "content": "<p>Thanks. I have used it in my kernel, but it's wrong when i train my model, i got auc with nan in the last 1000 mini-batch of every epoch, but everything else is ok, like loss, acc. </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "426407": "",
    "426719": "Area under the ROC curve between the predicted probability and the observed target.\n\nThis is mentioned on the [Overview Page ](https://www.kaggle.com/c/histopathologic-cancer-detection#evaluation) under __Evaluation__ Tab",
    "430841": "Thanks. I have used it in my kernel, but it's wrong when i train my model, i got auc with nan in the last 1000 mini-batch of every epoch, but everything else is ok, like loss, acc."
  },
  "source": "meta"
}