{
  "id": 340082,
  "title": "Baseline Model Question",
  "url": "/competitions/amex-default-prediction/discussion/340082",
  "author_name": "DW2022",
  "post_date": "2022-07-27T12:43:31.643000",
  "votes": 0,
  "comment_count": 8,
  "views": 0,
  "content": "<p>Is there any possibility that the baseline model could reach 0.800 in LB </p>",
  "messages": [
    {
      "id": 1873734,
      "postDate": "2022-07-27T21:42:00.517Z",
      "content": "<p>What is the \"baseline model\" you refer to?</p>",
      "rawMarkdown": "What is the \"baseline model\" you refer to?",
      "votes": 2,
      "replies": [
        {
          "id": 1873850,
          "postDate": "2022-07-27T23:46:54.083Z",
          "content": "<p>I am using lightGBM. 2 folds out of 5 in my local cv is really low (please see below) I am trying to figure it out why this happens. Is there any tips?</p>\n<p>fold 0 CV score is 0.8015<br>\n<strong>fold 1 CV score is 0.7913</strong><br>\nfold 2 CV score is 0.7973<br>\n<strong>fold 3 CV score is 0.7923</strong><br>\nfold 4 CV score is 0.79621</p>",
          "rawMarkdown": "I am using lightGBM. 2 folds out of 5 in my local cv is really low (please see below) I am trying to figure it out why this happens. Is there any tips?\n\nfold 0 CV score is 0.8015\n**fold 1 CV score is 0.7913**\nfold 2 CV score is 0.7973\n**fold 3 CV score is 0.7923**\nfold 4 CV score is 0.79621\n"
        },
        {
          "id": 1874353,
          "postDate": "2022-07-28T08:31:03.980Z",
          "content": "<p>I think fold 1 and 3 have more number of difficult predictions in their distribution. </p>",
          "rawMarkdown": "I think fold 1 and 3 have more number of difficult predictions in their distribution. "
        },
        {
          "id": 1876025,
          "postDate": "2022-07-29T14:19:33.207Z",
          "rawMarkdown": "",
          "isDeleted": true
        },
        {
          "id": 1877160,
          "postDate": "2022-07-30T11:47:55.743Z",
          "rawMarkdown": "",
          "isDeleted": true
        },
        {
          "id": 1877172,
          "postDate": "2022-07-30T11:52:51.310Z",
          "content": "<p>yes I use seed 42, I am doing some hyper parameter search to see if I could reach a little higher CV. </p>",
          "rawMarkdown": "yes I use seed 42, I am doing some hyper parameter search to see if I could reach a little higher CV. "
        }
      ]
    },
    {
      "id": 1873887,
      "postDate": "2022-07-28T00:27:59Z",
      "content": "<p>While I'm not 100% sure, I think it's definitely possible that the baseline model could reach 0.800 in the Leaderboard. I would encourage you to keep tweaking and testing your model to see if you can get it to that score.</p>\n<p>The Devastator.</p>",
      "rawMarkdown": "While I'm not 100% sure, I think it's definitely possible that the baseline model could reach 0.800 in the Leaderboard. I would encourage you to keep tweaking and testing your model to see if you can get it to that score.\n\n\nThe Devastator.\n"
    },
    {
      "id": 1873514,
      "postDate": "2022-07-27T16:40:37.480Z",
      "content": "<p>Yes, of course, it could happen. a baseline model is just another model! It could perhaps be the best option for the competition too!</p>",
      "rawMarkdown": "Yes, of course, it could happen. a baseline model is just another model! It could perhaps be the best option for the competition too!"
    },
    {
      "id": 1873084,
      "postDate": "2022-07-27T12:43:31.643Z",
      "content": "<p>Is there any possibility that the baseline model could reach 0.800 in LB </p>",
      "rawMarkdown": "Is there any possibility that the baseline model could reach 0.800 in LB "
    }
  ],
  "comments": [
    {
      "id": 1873734,
      "author_name": "Chris Deotte",
      "author_url": "",
      "post_date": "2022-07-27T21:42:00.517000",
      "content": "<p>What is the \"baseline model\" you refer to?</p>",
      "votes": 2,
      "replies": [
        {
          "id": 1873850,
          "author_name": "DW2022",
          "author_url": "",
          "post_date": "2022-07-27T23:46:54.083000",
          "content": "<p>I am using lightGBM. 2 folds out of 5 in my local cv is really low (please see below) I am trying to figure it out why this happens. Is there any tips?</p>\n<p>fold 0 CV score is 0.8015<br>\n<strong>fold 1 CV score is 0.7913</strong><br>\nfold 2 CV score is 0.7973<br>\n<strong>fold 3 CV score is 0.7923</strong><br>\nfold 4 CV score is 0.79621</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1874353,
          "author_name": "1110Ra",
          "author_url": "",
          "post_date": "2022-07-28T08:31:03.980000",
          "content": "<p>I think fold 1 and 3 have more number of difficult predictions in their distribution. </p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1876025,
          "author_name": "",
          "author_url": "",
          "post_date": "2022-07-29T14:19:33.207000",
          "content": "",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1877160,
          "author_name": "",
          "author_url": "",
          "post_date": "2022-07-30T11:47:55.743000",
          "content": "",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1877172,
          "author_name": "DW2022",
          "author_url": "",
          "post_date": "2022-07-30T11:52:51.310000",
          "content": "<p>yes I use seed 42, I am doing some hyper parameter search to see if I could reach a little higher CV. </p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 1873887,
      "author_name": "The Devastator",
      "author_url": "",
      "post_date": "2022-07-28T00:27:59",
      "content": "<p>While I'm not 100% sure, I think it's definitely possible that the baseline model could reach 0.800 in the Leaderboard. I would encourage you to keep tweaking and testing your model to see if you can get it to that score.</p>\n<p>The Devastator.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1873514,
      "author_name": "Ravi Ramakrishnan",
      "author_url": "",
      "post_date": "2022-07-27T16:40:37.480000",
      "content": "<p>Yes, of course, it could happen. a baseline model is just another model! It could perhaps be the best option for the competition too!</p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1873734": "What is the \"baseline model\" you refer to?",
    "1873887": "While I'm not 100% sure, I think it's definitely possible that the baseline model could reach 0.800 in the Leaderboard. I would encourage you to keep tweaking and testing your model to see if you can get it to that score.\n\n\nThe Devastator.\n",
    "1873514": "Yes, of course, it could happen. a baseline model is just another model! It could perhaps be the best option for the competition too!",
    "1873084": "Is there any possibility that the baseline model could reach 0.800 in LB "
  }
}