{
  "id": 507349,
  "title": "Imbalance target",
  "url": "/competitions/home-credit-credit-risk-model-stability/discussion/507349",
  "author_name": "",
  "post_date": "2024-05-25T10:56:17.635001800Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hello everyone,<br>\nI noticed that my final scores are consistently close to 0, with the highest only reaching 0.2. After further investigation, I found that the training data has a significant imbalance in the target variable: there are almost 1.4 million instances of 0 and only around 40,000 instances of 1. I suspect this imbalance is the reason for the low scores. </p>",
  "messages": [
    {
      "id": "2835508",
      "postDate": "05/25/2024 10:56:17",
      "content": "<p>Hello everyone,<br>\nI noticed that my final scores are consistently close to 0, with the highest only reaching 0.2. After further investigation, I found that the training data has a significant imbalance in the target variable: there are almost 1.4 million instances of 0 and only around 40,000 instances of 1. I suspect this imbalance is the reason for the low scores. </p>",
      "rawMarkdown": "Hello everyone,\n\nI noticed that my final scores are consistently close to 0, with the highest only reaching 0.2. After further investigation, I found that the training data has a significant imbalance in the target variable: there are almost 1.4 million instances of 0 and only around 40,000 instances of 1. I suspect this imbalance is the reason for the low scores.",
      "votes": null
    },
    {
      "id": "2835516",
      "postDate": "05/25/2024 10:59:35",
      "content": "<p>Hello, <a href=\"https://www.analyticsvidhya.com/blog/2020/07/10-techniques-to-deal-with-class-imbalance-in-machine-learning/\" target=\"_blank\">this website</a> might give you some ideas to solve your problem.</p>",
      "rawMarkdown": "Hello, [this website](https://www.analyticsvidhya.com/blog/2020/07/10-techniques-to-deal-with-class-imbalance-in-machine-learning/) might give you some ideas to solve your problem.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2835516,
      "author_name": "austinckr",
      "author_url": "",
      "post_date": "05/25/2024 10:59:35",
      "content": "<p>Hello, <a href=\"https://www.analyticsvidhya.com/blog/2020/07/10-techniques-to-deal-with-class-imbalance-in-machine-learning/\" target=\"_blank\">this website</a> might give you some ideas to solve your problem.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2835508": "Hello everyone,\n\nI noticed that my final scores are consistently close to 0, with the highest only reaching 0.2. After further investigation, I found that the training data has a significant imbalance in the target variable: there are almost 1.4 million instances of 0 and only around 40,000 instances of 1. I suspect this imbalance is the reason for the low scores.",
    "2835516": "Hello, [this website](https://www.analyticsvidhya.com/blog/2020/07/10-techniques-to-deal-with-class-imbalance-in-machine-learning/) might give you some ideas to solve your problem."
  },
  "source": "meta"
}