{
  "id": 407666,
  "title": "submission score is very low",
  "url": "/competitions/benetech-making-graphs-accessible/discussion/407666",
  "author_name": "",
  "post_date": "2023-05-07T14:14:49.934829300Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>When I submitted I have the low public score which is less than half of my evaluation score.<br>\nI evaluated my model by val dataset(a part of train data which has same chart type allocation as train data).<br>\nAnd use the following scoring method.</p>\n<p><a href=\"https://www.kaggle.com/code/ryanholbrook/competition-metric-benetech-mixed-match\" target=\"_blank\">https://www.kaggle.com/code/ryanholbrook/competition-metric-benetech-mixed-match</a><br>\nDo I have to modify above?</p>\n<p>Does anyone know the cause of this problem?</p>",
  "messages": [
    {
      "id": "2249172",
      "postDate": "05/07/2023 14:14:49",
      "content": "<p>When I submitted I have the low public score which is less than half of my evaluation score.<br>\nI evaluated my model by val dataset(a part of train data which has same chart type allocation as train data).<br>\nAnd use the following scoring method.</p>\n<p><a href=\"https://www.kaggle.com/code/ryanholbrook/competition-metric-benetech-mixed-match\" target=\"_blank\">https://www.kaggle.com/code/ryanholbrook/competition-metric-benetech-mixed-match</a><br>\nDo I have to modify above?</p>\n<p>Does anyone know the cause of this problem?</p>",
      "rawMarkdown": "When I submitted I have the low public score which is less than half of my evaluation score.\nI evaluated my model by val dataset(a part of train data which has same chart type allocation as train data).\nAnd use the following scoring method.\n\nhttps://www.kaggle.com/code/ryanholbrook/competition-metric-benetech-mixed-match\nDo I have to modify above?\n\nDoes anyone know the cause of this problem?",
      "votes": null
    },
    {
      "id": "2249561",
      "postDate": "05/07/2023 21:01:22",
      "content": "<p>Similar here. Not sure what's wrong.</p>",
      "rawMarkdown": "Similar here. Not sure what's wrong.",
      "votes": null
    },
    {
      "id": "2249609",
      "postDate": "05/08/2023 00:14:06",
      "content": "<p>I found this discussion just now.<br>\n<a href=\"https://www.kaggle.com/competitions/benetech-making-graphs-accessible/discussion/397635\" target=\"_blank\">https://www.kaggle.com/competitions/benetech-making-graphs-accessible/discussion/397635</a></p>\n<p>Test dataset have only \"extracted\" graphs and my validation score about only extracted dataset has similar to public score.<br>\nIt seems much more difficult to inference about \"extracted\" graphs than \"generated\" ones.</p>",
      "rawMarkdown": "I found this discussion just now.\nhttps://www.kaggle.com/competitions/benetech-making-graphs-accessible/discussion/397635\n\nTest dataset have only \"extracted\" graphs and my validation score about only extracted dataset has similar to public score.\nIt seems much more difficult to inference about \"extracted\" graphs than \"generated\" ones.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2249561,
      "author_name": "zhugds",
      "author_url": "",
      "post_date": "05/07/2023 21:01:22",
      "content": "<p>Similar here. Not sure what's wrong.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2249609,
      "author_name": "takafumifujii",
      "author_url": "",
      "post_date": "05/08/2023 00:14:06",
      "content": "<p>I found this discussion just now.<br>\n<a href=\"https://www.kaggle.com/competitions/benetech-making-graphs-accessible/discussion/397635\" target=\"_blank\">https://www.kaggle.com/competitions/benetech-making-graphs-accessible/discussion/397635</a></p>\n<p>Test dataset have only \"extracted\" graphs and my validation score about only extracted dataset has similar to public score.<br>\nIt seems much more difficult to inference about \"extracted\" graphs than \"generated\" ones.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2249172": "When I submitted I have the low public score which is less than half of my evaluation score.\nI evaluated my model by val dataset(a part of train data which has same chart type allocation as train data).\nAnd use the following scoring method.\n\nhttps://www.kaggle.com/code/ryanholbrook/competition-metric-benetech-mixed-match\nDo I have to modify above?\n\nDoes anyone know the cause of this problem?",
    "2249561": "Similar here. Not sure what's wrong.",
    "2249609": "I found this discussion just now.\nhttps://www.kaggle.com/competitions/benetech-making-graphs-accessible/discussion/397635\n\nTest dataset have only \"extracted\" graphs and my validation score about only extracted dataset has similar to public score.\nIt seems much more difficult to inference about \"extracted\" graphs than \"generated\" ones."
  },
  "source": "meta"
}