{
  "id": 609218,
  "title": "So, processing the upside-down data in the hideen dataset is the key point",
  "url": "/competitions/cmi-detect-behavior-with-sensor-data/discussion/609218",
  "author_name": "XuKong Ji",
  "post_date": "2025-09-25T02:35:46.916000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>As the topic, processing the upside-down data in the hidden dataset is the key point, especially for reducing the gap for pb score and private score. Again reminding me donn't bother to do EDA!!!</p>",
  "messages": [
    {
      "id": 3293918,
      "postDate": "2025-09-25T02:35:46.917Z",
      "content": "<p>As the topic, processing the upside-down data in the hidden dataset is the key point, especially for reducing the gap for pb score and private score. Again reminding me donn't bother to do EDA!!!</p>",
      "rawMarkdown": "As the topic, processing the upside-down data in the hidden dataset is the key point, especially for reducing the gap for pb score and private score. Again reminding me donn't bother to do EDA!!!",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "3293918": "As the topic, processing the upside-down data in the hidden dataset is the key point, especially for reducing the gap for pb score and private score. Again reminding me donn't bother to do EDA!!!"
  }
}