{
  "id": 6649,
  "title": "Galaxy A-Go-Go",
  "url": "/competitions/galaxy-zoo-the-galaxy-challenge/discussion/6649",
  "author_name": "",
  "post_date": "2013-12-20T23:40:17.930Z",
  "votes": 8,
  "comment_count": 6,
  "views": 4723,
  "content": "<p>The code for the Central Pixel Benchmark is available <a href=\"https://github.com/noahvanhoucke/GalaxyZooChallenge\">here</a>.&nbsp;For kicks, I've written it in <a href=\"http://golang.org/\">Go</a>.</p>",
  "messages": [
    {
      "id": "36432",
      "postDate": "12/20/2013 23:40:17",
      "content": "<p>The code for the Central Pixel Benchmark is available <a href=\"https://github.com/noahvanhoucke/GalaxyZooChallenge\">here</a>.&nbsp;For kicks, I've written it in <a href=\"http://golang.org/\">Go</a>.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "36441",
      "postDate": "12/21/2013 01:36:31",
      "content": "<p>This competition looks great, definitely gonna give it a go! I've been waiting for another image classification challenge and this is a nice change from classifying dogs, cats and handwritten digits :)</p>\n<p>I have a question regarding the evaluation metric, which seems to be the RMSE. If I'm not mistaken, the goal is essentially to predict a bunch of discrete probability distributions. Wouldn't it make more sense then to use something like a Kullback-Leibler divergence to evaluate how well the predictions match the true distributions?</p>\n<p>Was there a particular motivation to go with RMSE instead (e.g. speed), or am I perhaps misunderstanding something? Thanks!</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "36443",
      "postDate": "12/21/2013 02:52:57",
      "content": "<p>We considered KL but the Galaxy Zoo decision tree is not a &quot;proper&quot; tree so using KL would not have allowed the organizers to discriminate between the classes they were most interested in. &nbsp;The tree was designed with the citizen science effort in mind, not optimizing the machine learning effort we are currently undergoing.&nbsp;</p>\n<p>With the probability values assigned as described on the <a href=\"https://www.kaggle.com/c/galaxy-zoo-the-galaxy-challenge/details/the-galaxy-zoo-decision-tree\">GZ Decision Tree</a> page, RMSE is the most suitable metric.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "36445",
      "postDate": "12/21/2013 02:54:54",
      "content": "<p>Alright, thanks for the clarification!</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "36608",
      "postDate": "12/23/2013 23:52:28",
      "content": "<p>Oh wow, what led you to your choice of Go? I'm a fan of the language but usually use python or R for my statistics and data analysis work.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "36641",
      "postDate": "12/25/2013 16:18:46",
      "content": "<p>[quote=Bill DeRose;36608]</p>\n<p>Oh wow, what led you to your choice of Go? I'm a fan of the language but usually use python or R for my statistics and data analysis work.</p>\n<p>[/quote]</p>\n<p>This was my excuse to try the language out -- thankfully there is an Image library! I'd like to test out its concurrency on my next project (ran out of time with this one). I hear there is an NLP library but I'm not sure what else there is, so agree its usefulness for data work is a bit limited at the moment.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "36686",
      "postDate": "12/27/2013 04:06:41",
      "content": "<p>EDIT:&nbsp; 'kid b' beat me to it while I was composing.</p>\n<p>&nbsp;</p>\n<p>Thanks for the go starter code.&nbsp; To justify using the result file as a submission-tester, I figure I should at least be able to run it.</p>\n<p>for linux users, line 123 should be:</p>\n<p>galaxyId := strings.Split(listFiles[i], &quot;/&quot;)[N]</p>\n<p>difference being the foreslash ('/') and N where N is the index of the last list element when the path is split with '/'</p>\n<p>and for the go-newbies like myself, I found these links to be most helpful...</p>\n<p>http://phollow.fr/2012/11/building-installing-testing-golang/</p>\n<p>http://stackoverflow.com/questions/17524392/golang-error-when-i-try-do-install</p>\n<p>http://golang.org/doc/code.html</p>\n<p>&nbsp;</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 36441,
      "author_name": "sedielem",
      "author_url": "",
      "post_date": "12/21/2013 01:36:31",
      "content": "<p>This competition looks great, definitely gonna give it a go! I've been waiting for another image classification challenge and this is a nice change from classifying dogs, cats and handwritten digits :)</p>\n<p>I have a question regarding the evaluation metric, which seems to be the RMSE. If I'm not mistaken, the goal is essentially to predict a bunch of discrete probability distributions. Wouldn't it make more sense then to use something like a Kullback-Leibler divergence to evaluate how well the predictions match the true distributions?</p>\n<p>Was there a particular motivation to go with RMSE instead (e.g. speed), or am I perhaps misunderstanding something? Thanks!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 36443,
      "author_name": "joycenv",
      "author_url": "",
      "post_date": "12/21/2013 02:52:57",
      "content": "<p>We considered KL but the Galaxy Zoo decision tree is not a &quot;proper&quot; tree so using KL would not have allowed the organizers to discriminate between the classes they were most interested in. &nbsp;The tree was designed with the citizen science effort in mind, not optimizing the machine learning effort we are currently undergoing.&nbsp;</p>\n<p>With the probability values assigned as described on the <a href=\"https://www.kaggle.com/c/galaxy-zoo-the-galaxy-challenge/details/the-galaxy-zoo-decision-tree\">GZ Decision Tree</a> page, RMSE is the most suitable metric.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 36445,
      "author_name": "sedielem",
      "author_url": "",
      "post_date": "12/21/2013 02:54:54",
      "content": "<p>Alright, thanks for the clarification!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 36608,
      "author_name": "billderose",
      "author_url": "",
      "post_date": "12/23/2013 23:52:28",
      "content": "<p>Oh wow, what led you to your choice of Go? I'm a fan of the language but usually use python or R for my statistics and data analysis work.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 36641,
      "author_name": "joycenv",
      "author_url": "",
      "post_date": "12/25/2013 16:18:46",
      "content": "<p>[quote=Bill DeRose;36608]</p>\n<p>Oh wow, what led you to your choice of Go? I'm a fan of the language but usually use python or R for my statistics and data analysis work.</p>\n<p>[/quote]</p>\n<p>This was my excuse to try the language out -- thankfully there is an Image library! I'd like to test out its concurrency on my next project (ran out of time with this one). I hear there is an NLP library but I'm not sure what else there is, so agree its usefulness for data work is a bit limited at the moment.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 36686,
      "author_name": "zerozero",
      "author_url": "",
      "post_date": "12/27/2013 04:06:41",
      "content": "<p>EDIT:&nbsp; 'kid b' beat me to it while I was composing.</p>\n<p>&nbsp;</p>\n<p>Thanks for the go starter code.&nbsp; To justify using the result file as a submission-tester, I figure I should at least be able to run it.</p>\n<p>for linux users, line 123 should be:</p>\n<p>galaxyId := strings.Split(listFiles[i], &quot;/&quot;)[N]</p>\n<p>difference being the foreslash ('/') and N where N is the index of the last list element when the path is split with '/'</p>\n<p>and for the go-newbies like myself, I found these links to be most helpful...</p>\n<p>http://phollow.fr/2012/11/building-installing-testing-golang/</p>\n<p>http://stackoverflow.com/questions/17524392/golang-error-when-i-try-do-install</p>\n<p>http://golang.org/doc/code.html</p>\n<p>&nbsp;</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "36432": "",
    "36441": "",
    "36443": "",
    "36445": "",
    "36608": "",
    "36641": "",
    "36686": ""
  },
  "source": "meta"
}