{
  "id": 167563,
  "title": "Imbalanced datasets",
  "url": "/competitions/landmark-retrieval-2020/discussion/167563",
  "author_name": "JIN",
  "post_date": "2020-07-17T05:03:38.540000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi there !<br>\nWhat do you think about Imbalanced classes in Retrieval Challenge ?<br>\nOur team is trying to build a Large-scale Landmark datasets like a GLDv2. <br>\nTotal 5mn images &amp; 25k labels(200 images per labels).<br>\nI just want to discuss about it.</p>\n<p>Imbalanced datasets vs well distributed datasets.</p>\n<p>which one is better ? </p>",
  "messages": [
    {
      "id": 932486,
      "postDate": "2020-07-17T05:03:38.540Z",
      "content": "<p>Hi there !<br>\nWhat do you think about Imbalanced classes in Retrieval Challenge ?<br>\nOur team is trying to build a Large-scale Landmark datasets like a GLDv2. <br>\nTotal 5mn images &amp; 25k labels(200 images per labels).<br>\nI just want to discuss about it.</p>\n<p>Imbalanced datasets vs well distributed datasets.</p>\n<p>which one is better ? </p>",
      "rawMarkdown": "Hi there !\nWhat do you think about Imbalanced classes in Retrieval Challenge ?\nOur team is trying to build a Large-scale Landmark datasets like a GLDv2. \nTotal 5mn images &amp; 25k labels(200 images per labels).\nI just want to discuss about it.\n\nImbalanced datasets vs well distributed datasets.\n\nwhich one is better ? \n\n"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "932486": "Hi there !\nWhat do you think about Imbalanced classes in Retrieval Challenge ?\nOur team is trying to build a Large-scale Landmark datasets like a GLDv2. \nTotal 5mn images &amp; 25k labels(200 images per labels).\nI just want to discuss about it.\n\nImbalanced datasets vs well distributed datasets.\n\nwhich one is better ? \n\n"
  }
}