{
  "id": 338184,
  "title": "Why the lower the score, the better the ranking？",
  "url": "/competitions/mayo-clinic-strip-ai/discussion/338184",
  "author_name": "Rongsheng Wang",
  "post_date": "2022-07-19T14:21:08.394000",
  "votes": 7,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Hello，I found that the score of 0.4 is higher than that of 0.5. Why？</p>",
  "messages": [
    {
      "id": 1862214,
      "postDate": "2022-07-19T14:21:08.393Z",
      "content": "<p>Hello，I found that the score of 0.4 is higher than that of 0.5. Why？</p>",
      "rawMarkdown": "Hello，I found that the score of 0.4 is higher than that of 0.5. Why？",
      "votes": 7
    },
    {
      "id": 1863671,
      "postDate": "2022-07-20T12:35:23.853Z",
      "content": "<p>Hey <a href=\"https://www.kaggle.com/rongshengwang\" target=\"_blank\">@rongshengwang</a> , the metric for this competition is a \"loss\" function, the logarithmic loss, which means that it measure error (which you want to minimize) instead of accuracy (which you want to maximize). You can read more on the <a href=\"https://www.kaggle.com/competitions/mayo-clinic-strip-ai/overview/evaluation\" target=\"_blank\">Evaluation</a> page.</p>\n<p>Hope this answers your question!</p>",
      "rawMarkdown": "Hey @rongshengwang , the metric for this competition is a \"loss\" function, the logarithmic loss, which means that it measure error (which you want to minimize) instead of accuracy (which you want to maximize). You can read more on the [Evaluation](https://www.kaggle.com/competitions/mayo-clinic-strip-ai/overview/evaluation) page.\n\nHope this answers your question!",
      "votes": 4
    },
    {
      "id": 1869314,
      "postDate": "2022-07-24T16:57:31.417Z",
      "content": "<p>because we want to minimize the loss function </p>",
      "rawMarkdown": "because we want to minimize the loss function "
    },
    {
      "id": 1863370,
      "postDate": "2022-07-20T10:07:01.223Z",
      "content": "<p>Because we want to minimize the loss function.</p>",
      "rawMarkdown": "Because we want to minimize the loss function."
    },
    {
      "id": 1862831,
      "postDate": "2022-07-20T03:35:57.600Z",
      "content": "<p>Because the metric for this competition is weighted <strong>multi-class logarithmic loss</strong>. The closer the loss is to zero, the better.</p>",
      "rawMarkdown": "Because the metric for this competition is weighted **multi-class logarithmic loss**. The closer the loss is to zero, the better."
    },
    {
      "id": 1871803,
      "postDate": "2022-07-26T13:41:17.173Z",
      "content": "<p><a href=\"https://www.kaggle.com/ryanholbrook\" target=\"_blank\">@ryanholbrook</a> Thanks！</p>",
      "rawMarkdown": "@ryanholbrook Thanks！",
      "votes": 2
    }
  ],
  "comments": [
    {
      "id": 1863671,
      "author_name": "Ryan Holbrook",
      "author_url": "",
      "post_date": "2022-07-20T12:35:23.853000",
      "content": "<p>Hey <a href=\"https://www.kaggle.com/rongshengwang\" target=\"_blank\">@rongshengwang</a> , the metric for this competition is a \"loss\" function, the logarithmic loss, which means that it measure error (which you want to minimize) instead of accuracy (which you want to maximize). You can read more on the <a href=\"https://www.kaggle.com/competitions/mayo-clinic-strip-ai/overview/evaluation\" target=\"_blank\">Evaluation</a> page.</p>\n<p>Hope this answers your question!</p>",
      "votes": 4,
      "replies": []
    },
    {
      "id": 1869314,
      "author_name": "hamza jaouadi",
      "author_url": "",
      "post_date": "2022-07-24T16:57:31.417000",
      "content": "<p>because we want to minimize the loss function </p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1863370,
      "author_name": "pastafarian",
      "author_url": "",
      "post_date": "2022-07-20T10:07:01.223000",
      "content": "<p>Because we want to minimize the loss function.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1862831,
      "author_name": "David Roberts",
      "author_url": "",
      "post_date": "2022-07-20T03:35:57.600000",
      "content": "<p>Because the metric for this competition is weighted <strong>multi-class logarithmic loss</strong>. The closer the loss is to zero, the better.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1871803,
      "author_name": "Rongsheng Wang",
      "author_url": "",
      "post_date": "2022-07-26T13:41:17.173000",
      "content": "<p><a href=\"https://www.kaggle.com/ryanholbrook\" target=\"_blank\">@ryanholbrook</a> Thanks！</p>",
      "votes": 2,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1862214": "Hello，I found that the score of 0.4 is higher than that of 0.5. Why？",
    "1863671": "Hey @rongshengwang , the metric for this competition is a \"loss\" function, the logarithmic loss, which means that it measure error (which you want to minimize) instead of accuracy (which you want to maximize). You can read more on the [Evaluation](https://www.kaggle.com/competitions/mayo-clinic-strip-ai/overview/evaluation) page.\n\nHope this answers your question!",
    "1869314": "because we want to minimize the loss function ",
    "1863370": "Because we want to minimize the loss function.",
    "1862831": "Because the metric for this competition is weighted **multi-class logarithmic loss**. The closer the loss is to zero, the better.",
    "1871803": "@ryanholbrook Thanks！"
  }
}