{
  "id": 342835,
  "title": "Standarization and normalization",
  "url": "/competitions/mayo-clinic-strip-ai/discussion/342835",
  "author_name": "Himanshu Nayal",
  "post_date": "2022-08-09T01:35:19.126000",
  "votes": 0,
  "comment_count": 5,
  "views": 0,
  "content": "<p>If i use standarization in training and normalization(dividing by 255.0) in inference. Will it effect the performance by anyway ?</p>",
  "messages": [
    {
      "id": 1891534,
      "postDate": "2022-08-09T13:35:17.890Z",
      "content": "<p>It should not have any effect if used consistently.</p>",
      "rawMarkdown": "It should not have any effect if used consistently.",
      "votes": 1,
      "replies": [
        {
          "id": 1892196,
          "postDate": "2022-08-09T23:29:09.107Z",
          "content": "<p>ok. thanks</p>",
          "rawMarkdown": "ok. thanks"
        },
        {
          "id": 1893099,
          "postDate": "2022-08-10T14:25:23.137Z",
          "content": "<p>So jcerpent, to generalize, as long as the same transformations are applied on the data during training and inference, there should be no issues right ?</p>",
          "rawMarkdown": "So jcerpent, to generalize, as long as the same transformations are applied on the data during training and inference, there should be no issues right ?"
        },
        {
          "id": 1893119,
          "postDate": "2022-08-10T14:44:37.147Z",
          "content": "<p>From my experience, you're correct. Although in some cases there might be some minor deviations occuring; these should then be closely observed and interpreted by the designer.</p>",
          "rawMarkdown": "From my experience, you're correct. Although in some cases there might be some minor deviations occuring; these should then be closely observed and interpreted by the designer.",
          "votes": 2
        }
      ]
    },
    {
      "id": 1890691,
      "postDate": "2022-08-09T01:35:19.127Z",
      "content": "<p>If i use standarization in training and normalization(dividing by 255.0) in inference. Will it effect the performance by anyway ?</p>",
      "rawMarkdown": "If i use standarization in training and normalization(dividing by 255.0) in inference. Will it effect the performance by anyway ?"
    },
    {
      "id": 1891366,
      "postDate": "2022-08-09T11:33:55.130Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 1891534,
      "author_name": "jcerpent",
      "author_url": "",
      "post_date": "2022-08-09T13:35:17.890000",
      "content": "<p>It should not have any effect if used consistently.</p>",
      "votes": 1,
      "replies": [
        {
          "id": 1892196,
          "author_name": "Himanshu Nayal",
          "author_url": "",
          "post_date": "2022-08-09T23:29:09.107000",
          "content": "<p>ok. thanks</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1893099,
          "author_name": "Naga Sai Abhinay Devarinti",
          "author_url": "",
          "post_date": "2022-08-10T14:25:23.137000",
          "content": "<p>So jcerpent, to generalize, as long as the same transformations are applied on the data during training and inference, there should be no issues right ?</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1893119,
          "author_name": "jcerpent",
          "author_url": "",
          "post_date": "2022-08-10T14:44:37.147000",
          "content": "<p>From my experience, you're correct. Although in some cases there might be some minor deviations occuring; these should then be closely observed and interpreted by the designer.</p>",
          "votes": 2,
          "replies": []
        }
      ]
    },
    {
      "id": 1891366,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-08-09T11:33:55.130000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1891534": "It should not have any effect if used consistently.",
    "1890691": "If i use standarization in training and normalization(dividing by 255.0) in inference. Will it effect the performance by anyway ?",
    "1891366": ""
  }
}