{
  "id": 242977,
  "title": "Classification result",
  "url": "/competitions/gct-apac-d-and-a-race/discussion/242977",
  "author_name": "",
  "post_date": "2021-05-31T18:05:07.642491700Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>If we are training the dataset to binary classification result 0 and 1, ideally the submission grid also to be 0 and 1, given 0.5!!</p>",
  "messages": [
    {
      "id": "1330364",
      "postDate": "05/31/2021 18:05:07",
      "content": "<p>If we are training the dataset to binary classification result 0 and 1, ideally the submission grid also to be 0 and 1, given 0.5!!</p>",
      "rawMarkdown": "If we are training the dataset to binary classification result 0 and 1, ideally the submission grid also to be 0 and 1, given 0.5!!",
      "votes": null
    },
    {
      "id": "1331539",
      "postDate": "06/01/2021 14:26:25",
      "content": "<p>Yes the labels are 0 or 1. You can give your prediction in 0 or 1, or you can give your prediction as a probability of the label to be 1 (hence a number between 0 to 1). Note the measurement is AUC, so it might be more logical to have a probability prediction.</p>",
      "rawMarkdown": "Yes the labels are 0 or 1. You can give your prediction in 0 or 1, or you can give your prediction as a probability of the label to be 1 (hence a number between 0 to 1). Note the measurement is AUC, so it might be more logical to have a probability prediction.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1331539,
      "author_name": "nichaoku",
      "author_url": "",
      "post_date": "06/01/2021 14:26:25",
      "content": "<p>Yes the labels are 0 or 1. You can give your prediction in 0 or 1, or you can give your prediction as a probability of the label to be 1 (hence a number between 0 to 1). Note the measurement is AUC, so it might be more logical to have a probability prediction.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1330364": "If we are training the dataset to binary classification result 0 and 1, ideally the submission grid also to be 0 and 1, given 0.5!!",
    "1331539": "Yes the labels are 0 or 1. You can give your prediction in 0 or 1, or you can give your prediction as a probability of the label to be 1 (hence a number between 0 to 1). Note the measurement is AUC, so it might be more logical to have a probability prediction."
  },
  "source": "meta"
}