{
  "id": 71891,
  "title": "The number of labels for each image in the training set",
  "url": "/competitions/quickdraw-doodle-recognition/discussion/71891",
  "author_name": "XiaokangWang",
  "post_date": "2018-11-17T21:13:31.899000",
  "votes": -2,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Each image in the testing set is supposed to have 3 predicted labels. I saw there is only one label for each image in the training set. Is that true?</p>",
  "messages": [
    {
      "id": 423341,
      "postDate": "2018-11-18T01:00:45.613Z",
      "content": "<p>You need to train a model that predicts the category of each image. \nThe predictions of your model will give you the probability of an image to belong to a particular class. \nYou are expected to submit per image the 3 classes with higher probability, the so called top-k probability. In this case k=3</p>",
      "rawMarkdown": "You need to train a model that predicts the category of each image. \nThe predictions of your model will give you the probability of an image to belong to a particular class. \nYou are expected to submit per image the 3 classes with higher probability, the so called top-k probability. In this case k=3",
      "votes": 1
    },
    {
      "id": 423286,
      "postDate": "2018-11-17T21:13:31.900Z",
      "content": "<p>Each image in the testing set is supposed to have 3 predicted labels. I saw there is only one label for each image in the training set. Is that true?</p>",
      "rawMarkdown": "Each image in the testing set is supposed to have 3 predicted labels. I saw there is only one label for each image in the training set. Is that true?",
      "votes": -2
    }
  ],
  "comments": [
    {
      "id": 423341,
      "author_name": "Nuno Ferreira",
      "author_url": "",
      "post_date": "2018-11-18T01:00:45.613000",
      "content": "<p>You need to train a model that predicts the category of each image. \nThe predictions of your model will give you the probability of an image to belong to a particular class. \nYou are expected to submit per image the 3 classes with higher probability, the so called top-k probability. In this case k=3</p>",
      "votes": 1,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "423341": "You need to train a model that predicts the category of each image. \nThe predictions of your model will give you the probability of an image to belong to a particular class. \nYou are expected to submit per image the 3 classes with higher probability, the so called top-k probability. In this case k=3",
    "423286": "Each image in the testing set is supposed to have 3 predicted labels. I saw there is only one label for each image in the training set. Is that true?"
  }
}