{
  "id": 252641,
  "title": "Prediction with NN",
  "url": "/competitions/mlb-player-digital-engagement-forecasting/discussion/252641",
  "author_name": "",
  "post_date": "2021-07-13T08:51:41.833006500Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I have been trying to predict target variables with NN.<br>\nBut I'm suffering from memory error.<br>\nIf anyone who using NN, please tell me how to avoid this error.<br>\nEspecially, I want to hear how to use categorial variables.<br>\nI tried FeatuteHashing and Ordinal Encoder,but didn't improve.<br>\nI would greatly appreciate your response.</p>",
  "messages": [
    {
      "id": "1386111",
      "postDate": "07/13/2021 08:51:41",
      "content": "<p>I have been trying to predict target variables with NN.<br>\nBut I'm suffering from memory error.<br>\nIf anyone who using NN, please tell me how to avoid this error.<br>\nEspecially, I want to hear how to use categorial variables.<br>\nI tried FeatuteHashing and Ordinal Encoder,but didn't improve.<br>\nI would greatly appreciate your response.</p>",
      "rawMarkdown": "I have been trying to predict target variables with NN.\nBut I'm suffering from memory error.\nIf anyone who using NN, please tell me how to avoid this error.\nEspecially, I want to hear how to use categorial variables.\nI tried FeatuteHashing and Ordinal Encoder,but didn't improve.\nI would greatly appreciate your response.",
      "votes": null
    },
    {
      "id": "1386367",
      "postDate": "07/13/2021 12:49:01",
      "content": "<p>Memory is a real issue with this data set.  A couple of shared kernels have a memory reduction def - not sure which ones, but that helps.  The Events section is a biggie - if your not using that piece don't load.</p>",
      "rawMarkdown": "Memory is a real issue with this data set.  A couple of shared kernels have a memory reduction def - not sure which ones, but that helps.  The Events section is a biggie - if your not using that piece don't load.",
      "votes": null
    },
    {
      "id": "1387051",
      "postDate": "07/13/2021 22:54:29",
      "content": "<p>Can lower your batch size and epochs to deal with memory error. </p>",
      "rawMarkdown": "Can lower your batch size and epochs to deal with memory error.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1386367,
      "author_name": "pcjimmmy",
      "author_url": "",
      "post_date": "07/13/2021 12:49:01",
      "content": "<p>Memory is a real issue with this data set.  A couple of shared kernels have a memory reduction def - not sure which ones, but that helps.  The Events section is a biggie - if your not using that piece don't load.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1387051,
      "author_name": "toxicmaze",
      "author_url": "",
      "post_date": "07/13/2021 22:54:29",
      "content": "<p>Can lower your batch size and epochs to deal with memory error. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1386111": "I have been trying to predict target variables with NN.\nBut I'm suffering from memory error.\nIf anyone who using NN, please tell me how to avoid this error.\nEspecially, I want to hear how to use categorial variables.\nI tried FeatuteHashing and Ordinal Encoder,but didn't improve.\nI would greatly appreciate your response.",
    "1386367": "Memory is a real issue with this data set.  A couple of shared kernels have a memory reduction def - not sure which ones, but that helps.  The Events section is a biggie - if your not using that piece don't load.",
    "1387051": "Can lower your batch size and epochs to deal with memory error."
  },
  "source": "meta"
}