{
  "id": 576719,
  "title": "One-Out-Of-Fold CV Performance",
  "url": "/competitions/el-hackathon-2025/discussion/576719",
  "author_name": "",
  "post_date": "2025-05-06T18:09:05.956753700Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi everyone—do you know a more effective alternative to one‑out‑of‑fold CV? Feel free to post your CV scores below; here’s mine. Thanks!</p>\n<p>0 -&gt; 0.70<br>\n1 -&gt; 0.77<br>\n2-&gt; 0.634<br>\n3 -&gt; 0.646<br>\n4 -&gt; 0.653<br>\n5-&gt; 0.615</p>",
  "messages": [
    {
      "id": "3195174",
      "postDate": "05/06/2025 18:09:05",
      "content": "<p>Hi everyone—do you know a more effective alternative to one‑out‑of‑fold CV? Feel free to post your CV scores below; here’s mine. Thanks!</p>\n<p>0 -&gt; 0.70<br>\n1 -&gt; 0.77<br>\n2-&gt; 0.634<br>\n3 -&gt; 0.646<br>\n4 -&gt; 0.653<br>\n5-&gt; 0.615</p>",
      "rawMarkdown": "Hi everyone—do you know a more effective alternative to one‑out‑of‑fold CV? Feel free to post your CV scores below; here’s mine. Thanks!\n\n0 -> 0.70\n1 -> 0.77\n2-> 0.634\n3 -> 0.646\n4 -> 0.653\n5-> 0.615",
      "votes": null
    },
    {
      "id": "3196839",
      "postDate": "05/07/2025 13:40:33",
      "content": "<p>0 -&gt; 0.47<br>\n1 -&gt; 0.62</p>\n<p>The score fluctuates a lot when I change anything. Do you experience this? <a href=\"https://www.kaggle.com/jamalsaeedi\" target=\"_blank\">@jamalsaeedi</a> </p>",
      "rawMarkdown": "0 -> 0.47\n1 -> 0.62\n\nThe score fluctuates a lot when I change anything. Do you experience this? @jamalsaeedi",
      "votes": null
    },
    {
      "id": "3199138",
      "postDate": "05/10/2025 14:13:40",
      "content": "<p>I think accuracy could really depend on the preprocess you did. The preprocessing is the most important key in this task in my point of view. Do you want to team up <a href=\"https://www.kaggle.com/jamalsaeedi\" target=\"_blank\">@jamalsaeedi</a> and discuss about the thing that I tried?<br>\nHoever, I have the same score fluctuates like <a href=\"https://www.kaggle.com/jamalsaeedi\" target=\"_blank\">@jamalsaeedi</a>. I also try to see the image embedding. The S_1 and S_2 embedding seperate from others. This could be an evidance why the score will look like this.</p>",
      "rawMarkdown": "I think accuracy could really depend on the preprocess you did. The preprocessing is the most important key in this task in my point of view. Do you want to team up @jamalsaeedi and discuss about the thing that I tried?\nHoever, I have the same score fluctuates like @jamalsaeedi. I also try to see the image embedding. The S_1 and S_2 embedding seperate from others. This could be an evidance why the score will look like this.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3196839,
      "author_name": "ayakakirima",
      "author_url": "",
      "post_date": "05/07/2025 13:40:33",
      "content": "<p>0 -&gt; 0.47<br>\n1 -&gt; 0.62</p>\n<p>The score fluctuates a lot when I change anything. Do you experience this? <a href=\"https://www.kaggle.com/jamalsaeedi\" target=\"_blank\">@jamalsaeedi</a> </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3199138,
      "author_name": "dingyangwang",
      "author_url": "",
      "post_date": "05/10/2025 14:13:40",
      "content": "<p>I think accuracy could really depend on the preprocess you did. The preprocessing is the most important key in this task in my point of view. Do you want to team up <a href=\"https://www.kaggle.com/jamalsaeedi\" target=\"_blank\">@jamalsaeedi</a> and discuss about the thing that I tried?<br>\nHoever, I have the same score fluctuates like <a href=\"https://www.kaggle.com/jamalsaeedi\" target=\"_blank\">@jamalsaeedi</a>. I also try to see the image embedding. The S_1 and S_2 embedding seperate from others. This could be an evidance why the score will look like this.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3195174": "Hi everyone—do you know a more effective alternative to one‑out‑of‑fold CV? Feel free to post your CV scores below; here’s mine. Thanks!\n\n0 -> 0.70\n1 -> 0.77\n2-> 0.634\n3 -> 0.646\n4 -> 0.653\n5-> 0.615",
    "3196839": "0 -> 0.47\n1 -> 0.62\n\nThe score fluctuates a lot when I change anything. Do you experience this? @jamalsaeedi",
    "3199138": "I think accuracy could really depend on the preprocess you did. The preprocessing is the most important key in this task in my point of view. Do you want to team up @jamalsaeedi and discuss about the thing that I tried?\nHoever, I have the same score fluctuates like @jamalsaeedi. I also try to see the image embedding. The S_1 and S_2 embedding seperate from others. This could be an evidance why the score will look like this."
  },
  "source": "meta"
}