{
  "id": 74888,
  "title": "Why luminosity doesn't work ?",
  "url": "/competitions/PLAsTiCC-2018/discussion/74888",
  "author_name": "",
  "post_date": "2018-12-17T04:22:57.162748700Z",
  "votes": 2,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Hello everyone, </p>\n\n<p>I try to use <strong><em>Luminosity</em></strong>, this physical quantity is calculated by </p>\n\n<p>L = 4 x pi x distance**2 x flux</p>\n\n<p><em><strong>Luminosity</strong></em> is the total amount of energy emitted per unit of time by a star, galaxy, or other astronomical object, so I think it is useful for this competition. I compute distance with astropy,  luminosity_distance(hostgal_photoz).\nI add luminosity to some kernels, but my score is worse than original's.\nI have no idea why getting worse... If you have some knowledge, Please reply to this discussion topic. </p>\n\n<p>Thank you for reading.</p>",
  "messages": [
    {
      "id": "440116",
      "postDate": "12/17/2018 04:22:57",
      "content": "<p>Hello everyone, </p>\n\n<p>I try to use <strong><em>Luminosity</em></strong>, this physical quantity is calculated by </p>\n\n<p>L = 4 x pi x distance**2 x flux</p>\n\n<p><em><strong>Luminosity</strong></em> is the total amount of energy emitted per unit of time by a star, galaxy, or other astronomical object, so I think it is useful for this competition. I compute distance with astropy,  luminosity_distance(hostgal_photoz).\nI add luminosity to some kernels, but my score is worse than original's.\nI have no idea why getting worse... If you have some knowledge, Please reply to this discussion topic. </p>\n\n<p>Thank you for reading.</p>",
      "rawMarkdown": "Hello everyone, \n\nI try to use ***Luminosity***, this physical quantity is calculated by \n\nL = 4 x pi x distance**2 x flux\n\n***Luminosity*** is the total amount of energy emitted per unit of time by a star, galaxy, or other astronomical object, so I think it is useful for this competition. I compute distance with astropy,  luminosity_distance(hostgal_photoz).\nI add luminosity to some kernels, but my score is worse than original's.\nI have no idea why getting worse... If you have some knowledge, Please reply to this discussion topic. \n\nThank you for reading.",
      "votes": null
    },
    {
      "id": "440235",
      "postDate": "12/17/2018 09:13:00",
      "content": "<p>Unlike the flux, the luminosity is an intrinsic property of an astronomical body (think of it as the light content). The flux (or brightness) is how much of the energy radiating from the body is detected by an instrument (usually something like a bolometer). Since the luminosity does not vary for one particular object (over distance or time), perhaps you're adding redundant features? Not sure but it's a guess... </p>",
      "rawMarkdown": "Unlike the flux, the luminosity is an intrinsic property of an astronomical body (think of it as the light content). The flux (or brightness) is how much of the energy radiating from the body is detected by an instrument (usually something like a bolometer). Since the luminosity does not vary for one particular object (over distance or time), perhaps you're adding redundant features? Not sure but it's a guess...",
      "votes": null
    },
    {
      "id": "440253",
      "postDate": "12/17/2018 09:42:10",
      "content": "<p>You're right, I added unnecessary feature. It's possibly happened overfitting.\nSince distant astrophysical event's flux are is small, I thought it needs correction or something.\nBTW, thank you for your reply !</p>",
      "rawMarkdown": "You're right, I added unnecessary feature. It's possibly happened overfitting.\nSince distant astrophysical event's flux are is small, I thought it needs correction or something.\nBTW, thank you for your reply !",
      "votes": null
    },
    {
      "id": "440560",
      "postDate": "12/17/2018 17:45:45",
      "content": "<p>You're welcome! </p>",
      "rawMarkdown": "You're welcome!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 440235,
      "author_name": "delayedkarma",
      "author_url": "",
      "post_date": "12/17/2018 09:13:00",
      "content": "<p>Unlike the flux, the luminosity is an intrinsic property of an astronomical body (think of it as the light content). The flux (or brightness) is how much of the energy radiating from the body is detected by an instrument (usually something like a bolometer). Since the luminosity does not vary for one particular object (over distance or time), perhaps you're adding redundant features? Not sure but it's a guess... </p>",
      "votes": null,
      "replies": [
        {
          "id": 440253,
          "author_name": "nakasho",
          "author_url": "",
          "post_date": "12/17/2018 09:42:10",
          "content": "<p>You're right, I added unnecessary feature. It's possibly happened overfitting.\nSince distant astrophysical event's flux are is small, I thought it needs correction or something.\nBTW, thank you for your reply !</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 440560,
          "author_name": "delayedkarma",
          "author_url": "",
          "post_date": "12/17/2018 17:45:45",
          "content": "<p>You're welcome! </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "440116": "Hello everyone, \n\nI try to use ***Luminosity***, this physical quantity is calculated by \n\nL = 4 x pi x distance**2 x flux\n\n***Luminosity*** is the total amount of energy emitted per unit of time by a star, galaxy, or other astronomical object, so I think it is useful for this competition. I compute distance with astropy,  luminosity_distance(hostgal_photoz).\nI add luminosity to some kernels, but my score is worse than original's.\nI have no idea why getting worse... If you have some knowledge, Please reply to this discussion topic. \n\nThank you for reading.",
    "440235": "Unlike the flux, the luminosity is an intrinsic property of an astronomical body (think of it as the light content). The flux (or brightness) is how much of the energy radiating from the body is detected by an instrument (usually something like a bolometer). Since the luminosity does not vary for one particular object (over distance or time), perhaps you're adding redundant features? Not sure but it's a guess...",
    "440253": "You're right, I added unnecessary feature. It's possibly happened overfitting.\nSince distant astrophysical event's flux are is small, I thought it needs correction or something.\nBTW, thank you for your reply !",
    "440560": "You're welcome!"
  },
  "source": "meta"
}