{
  "id": 327263,
  "title": "What is the benchmark entry in the leaderboard?",
  "url": "/competitions/amex-default-prediction/discussion/327263",
  "author_name": "",
  "post_date": "2022-05-26T12:51:19.966592300Z",
  "votes": 9,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi,</p>\n<p>just taking a look at the leaderboard I saw this \"benchmark.csv\" entry. Can someone share what this is about? </p>",
  "messages": [
    {
      "id": "1802085",
      "postDate": "05/26/2022 12:51:19",
      "content": "<p>Hi,</p>\n<p>just taking a look at the leaderboard I saw this \"benchmark.csv\" entry. Can someone share what this is about? </p>",
      "rawMarkdown": "Hi,\n\njust taking a look at the leaderboard I saw this \"benchmark.csv\" entry. Can someone share what this is about?",
      "votes": null
    },
    {
      "id": "1802118",
      "postDate": "05/26/2022 13:19:38",
      "content": "<p>Most competitions will have a benchmark entry prepared by the competition organizers. Normally it's based on a very simple model, like picking the most common class in a classification problem or the mean of the target in a regression problem. In this competition there are two benchmark entries.</p>\n<ol>\n<li><strong>sample_submission.csv</strong> - the sample submission file included in the data set with a target prediction of all 0, which has a score of 0.019.</li>\n<li><strong>rf_benchmark.csv</strong> - based on the file name, predictions are probably based on a random forest model, which has a score of 0.751.</li>\n</ol>\n<p>These benchmark entries give us something to shoot for.</p>",
      "rawMarkdown": "Most competitions will have a benchmark entry prepared by the competition organizers. Normally it's based on a very simple model, like picking the most common class in a classification problem or the mean of the target in a regression problem. In this competition there are two benchmark entries.\n\n1. **sample_submission.csv** - the sample submission file included in the data set with a target prediction of all 0, which has a score of 0.019.\n2. **rf_benchmark.csv** - based on the file name, predictions are probably based on a random forest model, which has a score of 0.751.\n\nThese benchmark entries give us something to shoot for.",
      "votes": null
    },
    {
      "id": "1802270",
      "postDate": "05/26/2022 15:53:46",
      "content": "<p>Thanks a lot for the explanation</p>",
      "rawMarkdown": "Thanks a lot for the explanation",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1802118,
      "author_name": "bcruise",
      "author_url": "",
      "post_date": "05/26/2022 13:19:38",
      "content": "<p>Most competitions will have a benchmark entry prepared by the competition organizers. Normally it's based on a very simple model, like picking the most common class in a classification problem or the mean of the target in a regression problem. In this competition there are two benchmark entries.</p>\n<ol>\n<li><strong>sample_submission.csv</strong> - the sample submission file included in the data set with a target prediction of all 0, which has a score of 0.019.</li>\n<li><strong>rf_benchmark.csv</strong> - based on the file name, predictions are probably based on a random forest model, which has a score of 0.751.</li>\n</ol>\n<p>These benchmark entries give us something to shoot for.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1802270,
          "author_name": "thomasmeiner",
          "author_url": "",
          "post_date": "05/26/2022 15:53:46",
          "content": "<p>Thanks a lot for the explanation</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1802085": "Hi,\n\njust taking a look at the leaderboard I saw this \"benchmark.csv\" entry. Can someone share what this is about?",
    "1802118": "Most competitions will have a benchmark entry prepared by the competition organizers. Normally it's based on a very simple model, like picking the most common class in a classification problem or the mean of the target in a regression problem. In this competition there are two benchmark entries.\n\n1. **sample_submission.csv** - the sample submission file included in the data set with a target prediction of all 0, which has a score of 0.019.\n2. **rf_benchmark.csv** - based on the file name, predictions are probably based on a random forest model, which has a score of 0.751.\n\nThese benchmark entries give us something to shoot for.",
    "1802270": "Thanks a lot for the explanation"
  },
  "source": "meta"
}