{
  "id": 308766,
  "title": "The number of predictions to be made for a single image.",
  "url": "/competitions/happy-whale-and-dolphin/discussion/308766",
  "author_name": "",
  "post_date": "2022-02-20T07:57:29.126767600Z",
  "votes": 5,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hello.<br>\nI have a question about the CSV file to be submitted.<br>\nCan I have 5 predictions for one image?<br>\nI asked this question because when I looked at the 'sample csv' file, there were 5 predictions.</p>\n<p>Best regards.</p>",
  "messages": [
    {
      "id": "1698181",
      "postDate": "02/20/2022 07:57:29",
      "content": "<p>Hello.<br>\nI have a question about the CSV file to be submitted.<br>\nCan I have 5 predictions for one image?<br>\nI asked this question because when I looked at the 'sample csv' file, there were 5 predictions.</p>\n<p>Best regards.</p>",
      "rawMarkdown": "Hello.\nI have a question about the CSV file to be submitted.\nCan I have 5 predictions for one image?\nI asked this question because when I looked at the 'sample csv' file, there were 5 predictions.\n\nBest regards.",
      "votes": null
    },
    {
      "id": "1698334",
      "postDate": "02/20/2022 10:06:39",
      "content": "<p>Yes, you are supposed to give five predictions per single image. If your first prediction is correct it will bring you +1 score for the corresponding sample.<br>\nIf the second prediction is correct - a 1/2 score will be given.<br>\n…<br>\nThe last 5th prediction may give you only a 1/5 score.</p>\n<p>This is somehow similar to TopK accuracy, but additional weights are applied. It is an explanation in simple words. <br>\nThe more detailed description you may find in a <a href=\"https://www.kaggle.com/c/happy-whale-and-dolphin/overview/evaluation\" target=\"_blank\">evaluation page</a>.</p>\n<p>Also, there was an <a href=\"https://www.kaggle.com/pestipeti/explanation-of-map5-scoring-metric\" target=\"_blank\">amazing notebook with competition metric explanation</a> published by GM <a href=\"https://www.kaggle.com/pestipeti\" target=\"_blank\">@pestipeti</a> </p>",
      "rawMarkdown": "Yes, you are supposed to give five predictions per single image. If your first prediction is correct it will bring you +1 score for the corresponding sample.\nIf the second prediction is correct - a 1/2 score will be given.\n...\nThe last 5th prediction may give you only a 1/5 score.\n\nThis is somehow similar to TopK accuracy, but additional weights are applied. It is an explanation in simple words. \nThe more detailed description you may find in a [evaluation page](https://www.kaggle.com/c/happy-whale-and-dolphin/overview/evaluation).\n\nAlso, there was an [amazing notebook with competition metric explanation](https://www.kaggle.com/pestipeti/explanation-of-map5-scoring-metric) published by GM @pestipeti",
      "votes": null
    },
    {
      "id": "1698476",
      "postDate": "02/20/2022 12:09:19",
      "content": "<p>Hi Mykola!!<br>\nI was relieved with your advice.<br>\nThanks for nice advice.</p>",
      "rawMarkdown": "Hi Mykola!!\nI was relieved with your advice.\nThanks for nice advice.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1698334,
      "author_name": "meowmeowmeowmeowmeow",
      "author_url": "",
      "post_date": "02/20/2022 10:06:39",
      "content": "<p>Yes, you are supposed to give five predictions per single image. If your first prediction is correct it will bring you +1 score for the corresponding sample.<br>\nIf the second prediction is correct - a 1/2 score will be given.<br>\n…<br>\nThe last 5th prediction may give you only a 1/5 score.</p>\n<p>This is somehow similar to TopK accuracy, but additional weights are applied. It is an explanation in simple words. <br>\nThe more detailed description you may find in a <a href=\"https://www.kaggle.com/c/happy-whale-and-dolphin/overview/evaluation\" target=\"_blank\">evaluation page</a>.</p>\n<p>Also, there was an <a href=\"https://www.kaggle.com/pestipeti/explanation-of-map5-scoring-metric\" target=\"_blank\">amazing notebook with competition metric explanation</a> published by GM <a href=\"https://www.kaggle.com/pestipeti\" target=\"_blank\">@pestipeti</a> </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1698476,
      "author_name": "ishizakireo",
      "author_url": "",
      "post_date": "02/20/2022 12:09:19",
      "content": "<p>Hi Mykola!!<br>\nI was relieved with your advice.<br>\nThanks for nice advice.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1698181": "Hello.\nI have a question about the CSV file to be submitted.\nCan I have 5 predictions for one image?\nI asked this question because when I looked at the 'sample csv' file, there were 5 predictions.\n\nBest regards.",
    "1698334": "Yes, you are supposed to give five predictions per single image. If your first prediction is correct it will bring you +1 score for the corresponding sample.\nIf the second prediction is correct - a 1/2 score will be given.\n...\nThe last 5th prediction may give you only a 1/5 score.\n\nThis is somehow similar to TopK accuracy, but additional weights are applied. It is an explanation in simple words. \nThe more detailed description you may find in a [evaluation page](https://www.kaggle.com/c/happy-whale-and-dolphin/overview/evaluation).\n\nAlso, there was an [amazing notebook with competition metric explanation](https://www.kaggle.com/pestipeti/explanation-of-map5-scoring-metric) published by GM @pestipeti",
    "1698476": "Hi Mykola!!\nI was relieved with your advice.\nThanks for nice advice."
  },
  "source": "meta"
}