{
  "id": 177081,
  "title": "what is the sense in calculating the fvc for one observation? ",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/177081",
  "author_name": "",
  "post_date": "2020-08-24T19:05:09.667023700Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>what is the sense of calculating standard deviation of one value here FVC to population mean <br>\ndoesnt that mean FVC -( population FVCmean)?Or  dd they mean (Z*) *(standard deviation of whole population FVC)</p>",
  "messages": [
    {
      "id": "984011",
      "postDate": "08/24/2020 19:05:09",
      "content": "<p>what is the sense of calculating standard deviation of one value here FVC to population mean <br>\ndoesnt that mean FVC -( population FVCmean)?Or  dd they mean (Z*) *(standard deviation of whole population FVC)</p>",
      "rawMarkdown": "what is the sense of calculating standard deviation of one value here FVC to population mean \ndoesnt that mean FVC -( population FVCmean)?Or  dd they mean (Z*) *(standard deviation of whole population FVC)",
      "votes": null
    },
    {
      "id": "986405",
      "postDate": "08/26/2020 13:07:23",
      "content": "<p>Sorry, can you clarify the question? I am not sure I can understand it. You are asked to provide both the FVC prediction and standard deviation in your submission. Standard deviation here refers to the model confidence in its prediction, <strong><em>not</em></strong> the standard deviation of your predictions.</p>",
      "rawMarkdown": "Sorry, can you clarify the question? I am not sure I can understand it. You are asked to provide both the FVC prediction and standard deviation in your submission. Standard deviation here refers to the model confidence in its prediction, ***not*** the standard deviation of your predictions.",
      "votes": null
    },
    {
      "id": "988797",
      "postDate": "08/28/2020 09:34:23",
      "content": "<p><a href=\"https://www.kaggle.com/ahmedhshahin\" target=\"_blank\">@ahmedhshahin</a>  thankyou for the reply. <br>\nYes, I would clarify my question and hope you would understand my question better.<br>\nIsn't it that Standard deviation can be defined for the whole data set containing data points, I think it would be ambiguous to define the standard deviation of one data point in the dataset, right? But in the sample-submission.csv, the file contains the FCV for each week and standard deviation for each week separately in front of it. So how to calculate the standard deviation for each week separately as FCV value for each week is a data point in FCV distribution so the question again  breaks down to ' How we can calculate the standard deviation for each data point separately?'</p>\n<p>I do not understand what you meant to say by confidence 'in' Prediction but I would be extremely obliged if you give me give example or some explanation regarding the same? </p>",
      "rawMarkdown": "ahmedhshahin  thankyou for the reply. \nYes, I would clarify my question and hope you would understand my question better.\nIsn't it that Standard deviation can be defined for the whole data set containing data points, I think it would be ambiguous to define the standard deviation of one data point in the dataset, right? But in the sample-submission.csv, the file contains the FCV for each week and standard deviation for each week separately in front of it. So how to calculate the standard deviation for each week separately as FCV value for each week is a data point in FCV distribution so the question again  breaks down to ' How we can calculate the standard deviation for each data point separately?'\n\nI do not understand what you meant to say by confidence 'in' Prediction but I would be extremely obliged if you give me give example or some explanation regarding the same?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 986405,
      "author_name": "ahmedhshahin",
      "author_url": "",
      "post_date": "08/26/2020 13:07:23",
      "content": "<p>Sorry, can you clarify the question? I am not sure I can understand it. You are asked to provide both the FVC prediction and standard deviation in your submission. Standard deviation here refers to the model confidence in its prediction, <strong><em>not</em></strong> the standard deviation of your predictions.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 988797,
      "author_name": "amit12018",
      "author_url": "",
      "post_date": "08/28/2020 09:34:23",
      "content": "<p><a href=\"https://www.kaggle.com/ahmedhshahin\" target=\"_blank\">@ahmedhshahin</a>  thankyou for the reply. <br>\nYes, I would clarify my question and hope you would understand my question better.<br>\nIsn't it that Standard deviation can be defined for the whole data set containing data points, I think it would be ambiguous to define the standard deviation of one data point in the dataset, right? But in the sample-submission.csv, the file contains the FCV for each week and standard deviation for each week separately in front of it. So how to calculate the standard deviation for each week separately as FCV value for each week is a data point in FCV distribution so the question again  breaks down to ' How we can calculate the standard deviation for each data point separately?'</p>\n<p>I do not understand what you meant to say by confidence 'in' Prediction but I would be extremely obliged if you give me give example or some explanation regarding the same? </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "984011": "what is the sense of calculating standard deviation of one value here FVC to population mean \ndoesnt that mean FVC -( population FVCmean)?Or  dd they mean (Z*) *(standard deviation of whole population FVC)",
    "986405": "Sorry, can you clarify the question? I am not sure I can understand it. You are asked to provide both the FVC prediction and standard deviation in your submission. Standard deviation here refers to the model confidence in its prediction, ***not*** the standard deviation of your predictions.",
    "988797": "ahmedhshahin  thankyou for the reply. \nYes, I would clarify my question and hope you would understand my question better.\nIsn't it that Standard deviation can be defined for the whole data set containing data points, I think it would be ambiguous to define the standard deviation of one data point in the dataset, right? But in the sample-submission.csv, the file contains the FCV for each week and standard deviation for each week separately in front of it. So how to calculate the standard deviation for each week separately as FCV value for each week is a data point in FCV distribution so the question again  breaks down to ' How we can calculate the standard deviation for each data point separately?'\n\nI do not understand what you meant to say by confidence 'in' Prediction but I would be extremely obliged if you give me give example or some explanation regarding the same?"
  },
  "source": "meta"
}