{
  "id": 190060,
  "title": "Mean Baseline Submission sub['label'] = 0.2799 ?",
  "url": "/competitions/rsna-str-pulmonary-embolism-detection/discussion/190060",
  "author_name": "Jagadish Sivakumar",
  "post_date": "2020-10-10T00:55:16.980000",
  "votes": 3,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Hi Guys can somebody help me in understanding what <code>sub['label'] = 0.2799</code> refers to in the mean baseline submission kernels.</p>\n<p>And how many people conclude in choosing <code>0.2799</code> as the value. Several others have similar doubt, but there was no response in the respective kernel comment section.</p>\n<p>Thank you.</p>",
  "messages": [
    {
      "id": 1044716,
      "postDate": "2020-10-10T04:24:23.170Z",
      "content": "<p>The exam-level metric is logloss. The representative value minimizes logloss is average. On the other hand, the image-level metric is weighted logloss. The representative value minimizes weighted logloss is weighted average.<br>\nThe weighted average of pe_present_on_image is 0.289885. That’s why 0.2799 works good. </p>",
      "rawMarkdown": "The exam-level metric is logloss. The representative value minimizes logloss is average. On the other hand, the image-level metric is weighted logloss. The representative value minimizes weighted logloss is weighted average.\nThe weighted average of pe_present_on_image is 0.289885. That’s why 0.2799 works good. ",
      "votes": 4,
      "replies": [
        {
          "id": 1044763,
          "postDate": "2020-10-10T05:12:01.243Z",
          "content": "<p>Thanks <a href=\"https://www.kaggle.com/osciiart\" target=\"_blank\">@osciiart</a> 👍</p>",
          "rawMarkdown": "Thanks @osciiart 👍",
          "votes": 1
        }
      ]
    },
    {
      "id": 1044611,
      "postDate": "2020-10-10T00:55:16.980Z",
      "content": "<p>Hi Guys can somebody help me in understanding what <code>sub['label'] = 0.2799</code> refers to in the mean baseline submission kernels.</p>\n<p>And how many people conclude in choosing <code>0.2799</code> as the value. Several others have similar doubt, but there was no response in the respective kernel comment section.</p>\n<p>Thank you.</p>",
      "rawMarkdown": "Hi Guys can somebody help me in understanding what `sub['label'] = 0.2799` refers to in the mean baseline submission kernels.\n\nAnd how many people conclude in choosing `0.2799` as the value. Several others have similar doubt, but there was no response in the respective kernel comment section.\n\nThank you.",
      "votes": 2
    },
    {
      "id": 1044815,
      "postDate": "2020-10-10T06:26:50.167Z",
      "content": "<p>Brief idea on Log Loss:<br>\n<a href=\"https://www.kaggle.com/dansbecker/what-is-log-loss\" target=\"_blank\">https://www.kaggle.com/dansbecker/what-is-log-loss</a></p>",
      "rawMarkdown": "Brief idea on Log Loss:\nhttps://www.kaggle.com/dansbecker/what-is-log-loss",
      "votes": 2
    }
  ],
  "comments": [
    {
      "id": 1044716,
      "author_name": "OsciiArt",
      "author_url": "",
      "post_date": "2020-10-10T04:24:23.170000",
      "content": "<p>The exam-level metric is logloss. The representative value minimizes logloss is average. On the other hand, the image-level metric is weighted logloss. The representative value minimizes weighted logloss is weighted average.<br>\nThe weighted average of pe_present_on_image is 0.289885. That’s why 0.2799 works good. </p>",
      "votes": 4,
      "replies": [
        {
          "id": 1044763,
          "author_name": "Jagadish Sivakumar",
          "author_url": "",
          "post_date": "2020-10-10T05:12:01.243000",
          "content": "<p>Thanks <a href=\"https://www.kaggle.com/osciiart\" target=\"_blank\">@osciiart</a> 👍</p>",
          "votes": 1,
          "replies": []
        }
      ]
    },
    {
      "id": 1044815,
      "author_name": "Jagadish Sivakumar",
      "author_url": "",
      "post_date": "2020-10-10T06:26:50.167000",
      "content": "<p>Brief idea on Log Loss:<br>\n<a href=\"https://www.kaggle.com/dansbecker/what-is-log-loss\" target=\"_blank\">https://www.kaggle.com/dansbecker/what-is-log-loss</a></p>",
      "votes": 2,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1044716": "The exam-level metric is logloss. The representative value minimizes logloss is average. On the other hand, the image-level metric is weighted logloss. The representative value minimizes weighted logloss is weighted average.\nThe weighted average of pe_present_on_image is 0.289885. That’s why 0.2799 works good. ",
    "1044611": "Hi Guys can somebody help me in understanding what `sub['label'] = 0.2799` refers to in the mean baseline submission kernels.\n\nAnd how many people conclude in choosing `0.2799` as the value. Several others have similar doubt, but there was no response in the respective kernel comment section.\n\nThank you.",
    "1044815": "Brief idea on Log Loss:\nhttps://www.kaggle.com/dansbecker/what-is-log-loss"
  }
}