{
  "id": 482579,
  "title": "Intuition behind the frequencies selected by the organizer",
  "url": "/competitions/hms-harmful-brain-activity-classification/discussion/482579",
  "author_name": "",
  "post_date": "2024-03-08T10:13:29.290040Z",
  "votes": 6,
  "comment_count": 5,
  "views": 0,
  "content": "<p>I'm learning signal processing and I couldn't find anything regarding this. Are those values selected specifically for this task? Does anyone tried frequencies below 0.6 and above 20? Does it work for you?</p>",
  "messages": [
    {
      "id": "2687154",
      "postDate": "03/08/2024 10:13:29",
      "content": "<p>I'm learning signal processing and I couldn't find anything regarding this. Are those values selected specifically for this task? Does anyone tried frequencies below 0.6 and above 20? Does it work for you?</p>",
      "rawMarkdown": "I'm learning signal processing and I couldn't find anything regarding this. Are those values selected specifically for this task? Does anyone tried frequencies below 0.6 and above 20? Does it work for you?",
      "votes": null
    },
    {
      "id": "2687163",
      "postDate": "03/08/2024 10:26:13",
      "content": "<p>These are probably the default frequencies shown to experts for the task of classification of the data. So, the organizers just went with the same default parameters.</p>",
      "rawMarkdown": "These are probably the default frequencies shown to experts for the task of classification of the data. So, the organizers just went with the same default parameters.",
      "votes": null
    },
    {
      "id": "2695454",
      "postDate": "03/13/2024 17:28:47",
      "content": "<p>i think there so specific intuition behind the frequencies selected by the organizer …just default frequencies </p>",
      "rawMarkdown": "i think there so specific intuition behind the frequencies selected by the organizer ...just default frequencies",
      "votes": null
    },
    {
      "id": "2696977",
      "postDate": "03/14/2024 16:48:57",
      "content": "<p>I think it's EEG-specific. A few papers explaining this, but Wikipedia's explanation seems good enough: <a href=\"https://en.wikipedia.org/wiki/Electroencephalography\" target=\"_blank\">https://en.wikipedia.org/wiki/Electroencephalography</a></p>\n<blockquote>\n  <p>A healthy human EEG will show certain patterns of activity that correlate with how awake a person is. The range of frequencies one observes are between 1 and 30 Hz, and amplitudes will vary between 20 and 100 μV. The observed frequencies are subdivided into various groups: alpha (8–13 Hz), beta (13–30 Hz), delta (0.5–4 Hz), and theta (4–7 Hz). Alpha waves are observed when a person is in a state of relaxed wakefulness and are mostly prominent over the parietal and occipital sites.</p>\n</blockquote>\n<p>Typically for seizures, it seems frequencies above 20Hz aren't too relevant. Frequencies below 0.5Hz are below the delta range.</p>",
      "rawMarkdown": "I think it's EEG-specific. A few papers explaining this, but Wikipedia's explanation seems good enough: https://en.wikipedia.org/wiki/Electroencephalography\n\n> A healthy human EEG will show certain patterns of activity that correlate with how awake a person is. The range of frequencies one observes are between 1 and 30 Hz, and amplitudes will vary between 20 and 100 μV. The observed frequencies are subdivided into various groups: alpha (8–13 Hz), beta (13–30 Hz), delta (0.5–4 Hz), and theta (4–7 Hz). Alpha waves are observed when a person is in a state of relaxed wakefulness and are mostly prominent over the parietal and occipital sites.\n\nTypically for seizures, it seems frequencies above 20Hz aren't too relevant. Frequencies below 0.5Hz are below the delta range.",
      "votes": null
    },
    {
      "id": "2697010",
      "postDate": "03/14/2024 17:00:44",
      "content": "<p>Good find. I can also confirm that higher frequencies were useless in my models.</p>",
      "rawMarkdown": "Good find. I can also confirm that higher frequencies were useless in my models.",
      "votes": null
    },
    {
      "id": "2708963",
      "postDate": "03/21/2024 10:23:03",
      "content": "<p>Often in EEG signal analysis the lower frequencies are filtered out, with for example a high pass filter of 0.6Hz. This helps to normalize the signals as it will center the signal around zero and remove possible signal drift, see for example: <a href=\"url\" target=\"_blank\">https://benediktehinger.de/blog/science/electrode-drift-in-eeg/</a>. These drifts can be caused by an electrode coming loose or just by a natural shift in brain activation. </p>\n<p>Edit: The cut off frequency is often below 1Hz</p>",
      "rawMarkdown": "Often in EEG signal analysis the lower frequencies are filtered out, with for example a high pass filter of 0.6Hz. This helps to normalize the signals as it will center the signal around zero and remove possible signal drift, see for example: [https://benediktehinger.de/blog/science/electrode-drift-in-eeg/](url). These drifts can be caused by an electrode coming loose or just by a natural shift in brain activation. \n\nEdit: The cut off frequency is often below 1Hz",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2687163,
      "author_name": "shlomoron",
      "author_url": "",
      "post_date": "03/08/2024 10:26:13",
      "content": "<p>These are probably the default frequencies shown to experts for the task of classification of the data. So, the organizers just went with the same default parameters.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2695454,
      "author_name": "merugumalamanoj",
      "author_url": "",
      "post_date": "03/13/2024 17:28:47",
      "content": "<p>i think there so specific intuition behind the frequencies selected by the organizer …just default frequencies </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2696977,
      "author_name": "mandeepc",
      "author_url": "",
      "post_date": "03/14/2024 16:48:57",
      "content": "<p>I think it's EEG-specific. A few papers explaining this, but Wikipedia's explanation seems good enough: <a href=\"https://en.wikipedia.org/wiki/Electroencephalography\" target=\"_blank\">https://en.wikipedia.org/wiki/Electroencephalography</a></p>\n<blockquote>\n  <p>A healthy human EEG will show certain patterns of activity that correlate with how awake a person is. The range of frequencies one observes are between 1 and 30 Hz, and amplitudes will vary between 20 and 100 μV. The observed frequencies are subdivided into various groups: alpha (8–13 Hz), beta (13–30 Hz), delta (0.5–4 Hz), and theta (4–7 Hz). Alpha waves are observed when a person is in a state of relaxed wakefulness and are mostly prominent over the parietal and occipital sites.</p>\n</blockquote>\n<p>Typically for seizures, it seems frequencies above 20Hz aren't too relevant. Frequencies below 0.5Hz are below the delta range.</p>",
      "votes": null,
      "replies": [
        {
          "id": 2697010,
          "author_name": "gunesevitan",
          "author_url": "",
          "post_date": "03/14/2024 17:00:44",
          "content": "<p>Good find. I can also confirm that higher frequencies were useless in my models.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 2708963,
      "author_name": "matsrobben",
      "author_url": "",
      "post_date": "03/21/2024 10:23:03",
      "content": "<p>Often in EEG signal analysis the lower frequencies are filtered out, with for example a high pass filter of 0.6Hz. This helps to normalize the signals as it will center the signal around zero and remove possible signal drift, see for example: <a href=\"url\" target=\"_blank\">https://benediktehinger.de/blog/science/electrode-drift-in-eeg/</a>. These drifts can be caused by an electrode coming loose or just by a natural shift in brain activation. </p>\n<p>Edit: The cut off frequency is often below 1Hz</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2687154": "I'm learning signal processing and I couldn't find anything regarding this. Are those values selected specifically for this task? Does anyone tried frequencies below 0.6 and above 20? Does it work for you?",
    "2687163": "These are probably the default frequencies shown to experts for the task of classification of the data. So, the organizers just went with the same default parameters.",
    "2695454": "i think there so specific intuition behind the frequencies selected by the organizer ...just default frequencies",
    "2696977": "I think it's EEG-specific. A few papers explaining this, but Wikipedia's explanation seems good enough: https://en.wikipedia.org/wiki/Electroencephalography\n\n> A healthy human EEG will show certain patterns of activity that correlate with how awake a person is. The range of frequencies one observes are between 1 and 30 Hz, and amplitudes will vary between 20 and 100 μV. The observed frequencies are subdivided into various groups: alpha (8–13 Hz), beta (13–30 Hz), delta (0.5–4 Hz), and theta (4–7 Hz). Alpha waves are observed when a person is in a state of relaxed wakefulness and are mostly prominent over the parietal and occipital sites.\n\nTypically for seizures, it seems frequencies above 20Hz aren't too relevant. Frequencies below 0.5Hz are below the delta range.",
    "2697010": "Good find. I can also confirm that higher frequencies were useless in my models.",
    "2708963": "Often in EEG signal analysis the lower frequencies are filtered out, with for example a high pass filter of 0.6Hz. This helps to normalize the signals as it will center the signal around zero and remove possible signal drift, see for example: [https://benediktehinger.de/blog/science/electrode-drift-in-eeg/](url). These drifts can be caused by an electrode coming loose or just by a natural shift in brain activation. \n\nEdit: The cut off frequency is often below 1Hz"
  },
  "source": "meta"
}