{
  "id": 329736,
  "title": "Is there any chance of shake up?",
  "url": "/competitions/amex-default-prediction/discussion/329736",
  "author_name": "",
  "post_date": "2022-06-08T12:00:31.237324900Z",
  "votes": 8,
  "comment_count": 6,
  "views": 0,
  "content": "<p>For this competition is the cv and public lb correlated?</p>\n<p>I think the public lb is calculated on approximetely half of  the data so even if there is shakeup it would be minimum.<br>\nNeed other's opinion on this</p>",
  "messages": [
    {
      "id": "1814849",
      "postDate": "06/08/2022 12:00:31",
      "content": "<p>For this competition is the cv and public lb correlated?</p>\n<p>I think the public lb is calculated on approximetely half of  the data so even if there is shakeup it would be minimum.<br>\nNeed other's opinion on this</p>",
      "rawMarkdown": "For this competition is the cv and public lb correlated?\n\n I think the public lb is calculated on approximetely half of  the data so even if there is shakeup it would be minimum.\nNeed other's opinion on this",
      "votes": null
    },
    {
      "id": "1814871",
      "postDate": "06/08/2022 12:30:23",
      "content": "<p>Dear <a href=\"https://www.kaggle.com/yaswanthgali\" target=\"_blank\">@yaswanthgali</a> </p>\n<p>It seems we were both thinking about the shakeup, and posted topics on it at the same time! </p>\n<p>All the best,<br>\ncarl</p>",
      "rawMarkdown": "Dear @yaswanthgali \n\nIt seems we were both thinking about the shakeup, and posted topics on it at the same time! \n\nAll the best,\ncarl",
      "votes": null
    },
    {
      "id": "1814876",
      "postDate": "06/08/2022 12:35:41",
      "content": "<p>😂😂  i posted it coz i haven't made a good cross validation yet </p>",
      "rawMarkdown": "😂😂  i posted it coz i haven't made a good cross validation yet",
      "votes": null
    },
    {
      "id": "1814879",
      "postDate": "06/08/2022 12:40:02",
      "content": "<p>yep; I am also finding my hold-out test set scores are not a good reflection how well my model is really doing at all (different splitting seed gives different score)</p>",
      "rawMarkdown": "yep; I am also finding my hold-out test set scores are not a good reflection how well my model is really doing at all (different splitting seed gives different score)",
      "votes": null
    },
    {
      "id": "1815127",
      "postDate": "06/08/2022 17:35:00",
      "content": "<p>I also think it will be very small - the CV and LB results are almost identical, a large proportion of lines in public lb to private lb (especially taking into account recent competitions where there were 5%).</p>",
      "rawMarkdown": "I also think it will be very small - the CV and LB results are almost identical, a large proportion of lines in public lb to private lb (especially taking into account recent competitions where there were 5%).",
      "votes": null
    },
    {
      "id": "1815650",
      "postDate": "06/09/2022 09:08:25",
      "content": "<p>It will be really interesting to see how the time series element of this challenge impacts the results</p>",
      "rawMarkdown": "It will be really interesting to see how the time series element of this challenge impacts the results",
      "votes": null
    },
    {
      "id": "1815855",
      "postDate": "06/09/2022 14:49:22",
      "content": "<p>The problem is that a .002 shakeup would be relatively small … and the difference between no medal and gold right now!</p>",
      "rawMarkdown": "The problem is that a .002 shakeup would be relatively small ... and the difference between no medal and gold right now!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1814871,
      "author_name": "carlmcbrideellis",
      "author_url": "",
      "post_date": "06/08/2022 12:30:23",
      "content": "<p>Dear <a href=\"https://www.kaggle.com/yaswanthgali\" target=\"_blank\">@yaswanthgali</a> </p>\n<p>It seems we were both thinking about the shakeup, and posted topics on it at the same time! </p>\n<p>All the best,<br>\ncarl</p>",
      "votes": null,
      "replies": [
        {
          "id": 1814876,
          "author_name": "yaswanthgali",
          "author_url": "",
          "post_date": "06/08/2022 12:35:41",
          "content": "<p>😂😂  i posted it coz i haven't made a good cross validation yet </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1814879,
          "author_name": "carlmcbrideellis",
          "author_url": "",
          "post_date": "06/08/2022 12:40:02",
          "content": "<p>yep; I am also finding my hold-out test set scores are not a good reflection how well my model is really doing at all (different splitting seed gives different score)</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1815127,
      "author_name": "michau96",
      "author_url": "",
      "post_date": "06/08/2022 17:35:00",
      "content": "<p>I also think it will be very small - the CV and LB results are almost identical, a large proportion of lines in public lb to private lb (especially taking into account recent competitions where there were 5%).</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1815650,
      "author_name": "datajmcn",
      "author_url": "",
      "post_date": "06/09/2022 09:08:25",
      "content": "<p>It will be really interesting to see how the time series element of this challenge impacts the results</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1815855,
      "author_name": "happycube",
      "author_url": "",
      "post_date": "06/09/2022 14:49:22",
      "content": "<p>The problem is that a .002 shakeup would be relatively small … and the difference between no medal and gold right now!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1814849": "For this competition is the cv and public lb correlated?\n\n I think the public lb is calculated on approximetely half of  the data so even if there is shakeup it would be minimum.\nNeed other's opinion on this",
    "1814871": "Dear @yaswanthgali \n\nIt seems we were both thinking about the shakeup, and posted topics on it at the same time! \n\nAll the best,\ncarl",
    "1814876": "😂😂  i posted it coz i haven't made a good cross validation yet",
    "1814879": "yep; I am also finding my hold-out test set scores are not a good reflection how well my model is really doing at all (different splitting seed gives different score)",
    "1815127": "I also think it will be very small - the CV and LB results are almost identical, a large proportion of lines in public lb to private lb (especially taking into account recent competitions where there were 5%).",
    "1815650": "It will be really interesting to see how the time series element of this challenge impacts the results",
    "1815855": "The problem is that a .002 shakeup would be relatively small ... and the difference between no medal and gold right now!"
  },
  "source": "meta"
}