{
  "id": 305709,
  "title": "What do we need to predict?",
  "url": "/competitions/happy-whale-and-dolphin/discussion/305709",
  "author_name": "",
  "post_date": "2022-02-06T15:06:28.596360700Z",
  "votes": 3,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I know my question may sound silly but I am quite confused here. As per my understanding, given a test image we have to predict whether that image is that of an individual spotted in the train image or if it was never captured earlier then predict as \"new_individual\". So, this means that the submission file should either have a predicted individual (id) or a \"new_individual\" but why then the sample submission file has more then one entry in the \"predictions\" column for each of the test images.</p>",
  "messages": [
    {
      "id": "1678487",
      "postDate": "02/06/2022 15:06:28",
      "content": "<p>I know my question may sound silly but I am quite confused here. As per my understanding, given a test image we have to predict whether that image is that of an individual spotted in the train image or if it was never captured earlier then predict as \"new_individual\". So, this means that the submission file should either have a predicted individual (id) or a \"new_individual\" but why then the sample submission file has more then one entry in the \"predictions\" column for each of the test images.</p>",
      "rawMarkdown": "I know my question may sound silly but I am quite confused here. As per my understanding, given a test image we have to predict whether that image is that of an individual spotted in the train image or if it was never captured earlier then predict as \"new_individual\". So, this means that the submission file should either have a predicted individual (id) or a \"new_individual\" but why then the sample submission file has more then one entry in the \"predictions\" column for each of the test images.",
      "votes": null
    },
    {
      "id": "1678517",
      "postDate": "02/06/2022 15:29:14",
      "content": "<p>For each test image you may predict up to 5 individuals like \"I'm not sure who exactly but one of this five\" see competition metric for details</p>",
      "rawMarkdown": "For each test image you may predict up to 5 individuals like \"I'm not sure who exactly but one of this five\" see competition metric for details",
      "votes": null
    },
    {
      "id": "1678528",
      "postDate": "02/06/2022 15:37:30",
      "content": "<p>Thanks <a href=\"https://www.kaggle.com/kwentar\" target=\"_blank\">@kwentar</a> now it's clear . I totally missed this part. </p>",
      "rawMarkdown": "Thanks @kwentar now it's clear . I totally missed this part.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1678517,
      "author_name": "kwentar",
      "author_url": "",
      "post_date": "02/06/2022 15:29:14",
      "content": "<p>For each test image you may predict up to 5 individuals like \"I'm not sure who exactly but one of this five\" see competition metric for details</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1678528,
      "author_name": "sapal6",
      "author_url": "",
      "post_date": "02/06/2022 15:37:30",
      "content": "<p>Thanks <a href=\"https://www.kaggle.com/kwentar\" target=\"_blank\">@kwentar</a> now it's clear . I totally missed this part. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1678487": "I know my question may sound silly but I am quite confused here. As per my understanding, given a test image we have to predict whether that image is that of an individual spotted in the train image or if it was never captured earlier then predict as \"new_individual\". So, this means that the submission file should either have a predicted individual (id) or a \"new_individual\" but why then the sample submission file has more then one entry in the \"predictions\" column for each of the test images.",
    "1678517": "For each test image you may predict up to 5 individuals like \"I'm not sure who exactly but one of this five\" see competition metric for details",
    "1678528": "Thanks @kwentar now it's clear . I totally missed this part."
  },
  "source": "meta"
}