{
  "id": 82032,
  "title": "Proper mode of tackling this problem?",
  "url": "/competitions/LANL-Earthquake-Prediction/discussion/82032",
  "author_name": "",
  "post_date": "2019-02-26T23:15:10.141088500Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>In general, I am not a novice to machine learning or statistics ˆˆ, but I am still mostly entering this competition to get my feet wet with real time series forecasting.</p>\n\n<p>However, on that aspect I would appreciate a few pointers. Am I right to conclude that Survival Analysis is the main way to go about this competition? I say that because creating a time-to-event model (which Survival analysis seems to do) seems most fit to this question.</p>\n\n<p>I would greatly appreciate any advice.</p>",
  "messages": [
    {
      "id": "479015",
      "postDate": "02/26/2019 23:15:10",
      "content": "<p>In general, I am not a novice to machine learning or statistics ˆˆ, but I am still mostly entering this competition to get my feet wet with real time series forecasting.</p>\n\n<p>However, on that aspect I would appreciate a few pointers. Am I right to conclude that Survival Analysis is the main way to go about this competition? I say that because creating a time-to-event model (which Survival analysis seems to do) seems most fit to this question.</p>\n\n<p>I would greatly appreciate any advice.</p>",
      "rawMarkdown": "In general, I am not a novice to machine learning or statistics ˆˆ, but I am still mostly entering this competition to get my feet wet with real time series forecasting.\n\nHowever, on that aspect I would appreciate a few pointers. Am I right to conclude that Survival Analysis is the main way to go about this competition? I say that because creating a time-to-event model (which Survival analysis seems to do) seems most fit to this question.\n\nI would greatly appreciate any advice.",
      "votes": null
    },
    {
      "id": "480613",
      "postDate": "02/28/2019 12:24:07",
      "content": "<p>Yep, you're right, that's survival analysis as it is - you need to find statistical approach to investigate the time it takes for an event of interest  (an earthquake in our case) to occur.</p>",
      "rawMarkdown": "Yep, you're right, that's survival analysis as it is - you need to find statistical approach to investigate the time it takes for an event of interest  (an earthquake in our case) to occur.",
      "votes": null
    },
    {
      "id": "485664",
      "postDate": "03/07/2019 18:43:21",
      "content": "<p>Hi.\nI'm very newbie with survival analysis. I'm just read about this days\nAnd as you say it seems to be a very good way to fit the problem of predicting eartquake before it occurs (time to event).\nBut the big problem here is that when we predict a \"time to failure\" (y) giving a test segment (X) to a trained model , it is impossible to know what it arrived before the segment because the  13% of the public test set was shuffled. But survival analysis assumes that you know the past. No?</p>",
      "rawMarkdown": "Hi.\nI'm very newbie with survival analysis. I'm just read about this days\nAnd as you say it seems to be a very good way to fit the problem of predicting eartquake before it occurs (time to event).\nBut the big problem here is that when we predict a \"time to failure\" (y) giving a test segment (X) to a trained model , it is impossible to know what it arrived before the segment because the  13% of the public test set was shuffled. But survival analysis assumes that you know the past. No?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 480613,
      "author_name": "stanislavblinov",
      "author_url": "",
      "post_date": "02/28/2019 12:24:07",
      "content": "<p>Yep, you're right, that's survival analysis as it is - you need to find statistical approach to investigate the time it takes for an event of interest  (an earthquake in our case) to occur.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 485664,
      "author_name": "crunchyfanch",
      "author_url": "",
      "post_date": "03/07/2019 18:43:21",
      "content": "<p>Hi.\nI'm very newbie with survival analysis. I'm just read about this days\nAnd as you say it seems to be a very good way to fit the problem of predicting eartquake before it occurs (time to event).\nBut the big problem here is that when we predict a \"time to failure\" (y) giving a test segment (X) to a trained model , it is impossible to know what it arrived before the segment because the  13% of the public test set was shuffled. But survival analysis assumes that you know the past. No?</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "479015": "In general, I am not a novice to machine learning or statistics ˆˆ, but I am still mostly entering this competition to get my feet wet with real time series forecasting.\n\nHowever, on that aspect I would appreciate a few pointers. Am I right to conclude that Survival Analysis is the main way to go about this competition? I say that because creating a time-to-event model (which Survival analysis seems to do) seems most fit to this question.\n\nI would greatly appreciate any advice.",
    "480613": "Yep, you're right, that's survival analysis as it is - you need to find statistical approach to investigate the time it takes for an event of interest  (an earthquake in our case) to occur.",
    "485664": "Hi.\nI'm very newbie with survival analysis. I'm just read about this days\nAnd as you say it seems to be a very good way to fit the problem of predicting eartquake before it occurs (time to event).\nBut the big problem here is that when we predict a \"time to failure\" (y) giving a test segment (X) to a trained model , it is impossible to know what it arrived before the segment because the  13% of the public test set was shuffled. But survival analysis assumes that you know the past. No?"
  },
  "source": "meta"
}