{
  "id": 473148,
  "title": "Timeframe considerations for EEG spectrograms",
  "url": "/competitions/hms-harmful-brain-activity-classification/discussion/473148",
  "author_name": "Andreas Bisiadis",
  "post_date": "2024-02-03T17:45:50.083000",
  "votes": 4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hello there,</p>\n<p>I have trained EfficientNetB0 and ResNet34d on the EEG spectrograms using the full data (10 minutes or 600 seconds). Should I keep using the full timeframe or narrow it down to a specific duration (300 seconds)?</p>",
  "messages": [
    {
      "id": 2634348,
      "postDate": "2024-02-03T17:45:50.083Z",
      "content": "<p>Hello there,</p>\n<p>I have trained EfficientNetB0 and ResNet34d on the EEG spectrograms using the full data (10 minutes or 600 seconds). Should I keep using the full timeframe or narrow it down to a specific duration (300 seconds)?</p>",
      "rawMarkdown": "Hello there,\n\nI have trained EfficientNetB0 and ResNet34d on the EEG spectrograms using the full data (10 minutes or 600 seconds). Should I keep using the full timeframe or narrow it down to a specific duration (300 seconds)?",
      "votes": 4
    },
    {
      "id": 2635738,
      "postDate": "2024-02-04T15:23:10.093Z",
      "content": "<p>My <a href=\"https://www.kaggle.com/code/pcjimmmy/patient-variation-eda\" target=\"_blank\">EDA</a> suggests that some of the targets may have a lot of variation during an eeg.   (or not much difference in variation, just harder for evaluators to see and rate).</p>\n<p>Try 300 and find out is always good advice - just watch how you handle target values.</p>",
      "rawMarkdown": "My [EDA](https://www.kaggle.com/code/pcjimmmy/patient-variation-eda) suggests that some of the targets may have a lot of variation during an eeg.   (or not much difference in variation, just harder for evaluators to see and rate).\n\nTry 300 and find out is always good advice - just watch how you handle target values.\n\n",
      "votes": 1,
      "replies": [
        {
          "id": 2641831,
          "postDate": "2024-02-07T17:38:03.343Z",
          "rawMarkdown": "",
          "isDeleted": true
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 2635738,
      "author_name": "PC Jimmmy",
      "author_url": "",
      "post_date": "2024-02-04T15:23:10.093000",
      "content": "<p>My <a href=\"https://www.kaggle.com/code/pcjimmmy/patient-variation-eda\" target=\"_blank\">EDA</a> suggests that some of the targets may have a lot of variation during an eeg.   (or not much difference in variation, just harder for evaluators to see and rate).</p>\n<p>Try 300 and find out is always good advice - just watch how you handle target values.</p>",
      "votes": 1,
      "replies": [
        {
          "id": 2641831,
          "author_name": "",
          "author_url": "",
          "post_date": "2024-02-07T17:38:03.343000",
          "content": "",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2634348": "Hello there,\n\nI have trained EfficientNetB0 and ResNet34d on the EEG spectrograms using the full data (10 minutes or 600 seconds). Should I keep using the full timeframe or narrow it down to a specific duration (300 seconds)?",
    "2635738": "My [EDA](https://www.kaggle.com/code/pcjimmmy/patient-variation-eda) suggests that some of the targets may have a lot of variation during an eeg.   (or not much difference in variation, just harder for evaluators to see and rate).\n\nTry 300 and find out is always good advice - just watch how you handle target values.\n\n"
  }
}