{
  "id": 305259,
  "title": "Problem understanding \"Prediction\" column in Submission file",
  "url": "/competitions/happy-whale-and-dolphin/discussion/305259",
  "author_name": "",
  "post_date": "2022-02-04T13:15:51.156579900Z",
  "votes": 4,
  "comment_count": 1,
  "views": 0,
  "content": "<p>In each row of the submission file we have an image, and each image has a whale/dolphin, but in the prediction column we have several individual_id, shouldn't we just have one individual_id?</p>",
  "messages": [
    {
      "id": "1675771",
      "postDate": "02/04/2022 13:15:51",
      "content": "<p>In each row of the submission file we have an image, and each image has a whale/dolphin, but in the prediction column we have several individual_id, shouldn't we just have one individual_id?</p>",
      "rawMarkdown": "In each row of the submission file we have an image, and each image has a whale/dolphin, but in the prediction column we have several individual_id, shouldn't we just have one individual_id?",
      "votes": null
    },
    {
      "id": "1675874",
      "postDate": "02/04/2022 14:06:16",
      "content": "<p>In this competition submissions are evaluated according to the Mean Average Precision <a href=\"https://www.kaggle.com/5\" target=\"_blank\">@5</a> (MAP@5): For each image in the test set, you may predict up to 5 individual_id labels. </p>\n<p>It means you have up to five attempts to predict individuals correctly (Top 5 predictions). Note, that giving correct predictions at ranks #1 and #5 is not the same in terms of score. </p>\n<p>Check out <a href=\"https://www.kaggle.com/c/happy-whale-and-dolphin/overview/evaluation\" target=\"_blank\">the evaluation metrics page</a> for more details</p>",
      "rawMarkdown": "In this competition submissions are evaluated according to the Mean Average Precision @5 (MAP@5): For each image in the test set, you may predict up to 5 individual_id labels. \n\nIt means you have up to five attempts to predict individuals correctly (Top 5 predictions). Note, that giving correct predictions at ranks #1 and #5 is not the same in terms of score. \n\nCheck out [the evaluation metrics page](https://www.kaggle.com/c/happy-whale-and-dolphin/overview/evaluation) for more details",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1675874,
      "author_name": "meowmeowmeowmeowmeow",
      "author_url": "",
      "post_date": "02/04/2022 14:06:16",
      "content": "<p>In this competition submissions are evaluated according to the Mean Average Precision <a href=\"https://www.kaggle.com/5\" target=\"_blank\">@5</a> (MAP@5): For each image in the test set, you may predict up to 5 individual_id labels. </p>\n<p>It means you have up to five attempts to predict individuals correctly (Top 5 predictions). Note, that giving correct predictions at ranks #1 and #5 is not the same in terms of score. </p>\n<p>Check out <a href=\"https://www.kaggle.com/c/happy-whale-and-dolphin/overview/evaluation\" target=\"_blank\">the evaluation metrics page</a> for more details</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1675771": "In each row of the submission file we have an image, and each image has a whale/dolphin, but in the prediction column we have several individual_id, shouldn't we just have one individual_id?",
    "1675874": "In this competition submissions are evaluated according to the Mean Average Precision @5 (MAP@5): For each image in the test set, you may predict up to 5 individual_id labels. \n\nIt means you have up to five attempts to predict individuals correctly (Top 5 predictions). Note, that giving correct predictions at ranks #1 and #5 is not the same in terms of score. \n\nCheck out [the evaluation metrics page](https://www.kaggle.com/c/happy-whale-and-dolphin/overview/evaluation) for more details"
  },
  "source": "meta"
}