{
  "id": 23629,
  "title": "clarification on Cat dog competition?",
  "url": "/competitions/dogs-vs-cats-redux-kernels-edition/discussion/23629",
  "author_name": "",
  "post_date": "2016-09-13T20:16:12.323Z",
  "votes": null,
  "comment_count": 1,
  "views": 403,
  "content": "<p>Hey, I am new to kaggle and it's rules. I want to know if I submit a csv file, how will anyone judge that it's not manually overridden. \nSuppose I use caffe for prediction, do I need to disclose my dataset ever or it's just the model files I need to disclose?</p>",
  "messages": [
    {
      "id": "135400",
      "postDate": "09/13/2016 20:16:12",
      "content": "<p>Hey, I am new to kaggle and it's rules. I want to know if I submit a csv file, how will anyone judge that it's not manually overridden. \nSuppose I use caffe for prediction, do I need to disclose my dataset ever or it's just the model files I need to disclose?</p>",
      "rawMarkdown": "Hey, I am new to kaggle and it's rules. I want to know if I submit a csv file, how will anyone judge that it's not manually overridden. \r\nSuppose I use caffe for prediction, do I need to disclose my dataset ever or it's just the model files I need to disclose?",
      "votes": null
    },
    {
      "id": "135467",
      "postDate": "09/14/2016 10:42:33",
      "content": "<p>Hi, @installtt , no one will judge you.  Currently Leaderboard(LB) score is calculated on the first LB dataset, but then eventually final score will be evaluated on the other private dataset. This is why overfitting is always upon you. As I know this is just fun competition, so one will ask you for the model, but if you win payable competition, you  will need to disclose your model.</p>",
      "rawMarkdown": "Hi, @installtt , no one will judge you.  Currently Leaderboard(LB) score is calculated on the first LB dataset, but then eventually final score will be evaluated on the other private dataset. This is why overfitting is always upon you. As I know this is just fun competition, so one will ask you for the model, but if you win payable competition, you  will need to disclose your model.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 135467,
      "author_name": "insaff",
      "author_url": "",
      "post_date": "09/14/2016 10:42:33",
      "content": "<p>Hi, @installtt , no one will judge you.  Currently Leaderboard(LB) score is calculated on the first LB dataset, but then eventually final score will be evaluated on the other private dataset. This is why overfitting is always upon you. As I know this is just fun competition, so one will ask you for the model, but if you win payable competition, you  will need to disclose your model.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "135400": "Hey, I am new to kaggle and it's rules. I want to know if I submit a csv file, how will anyone judge that it's not manually overridden. \r\nSuppose I use caffe for prediction, do I need to disclose my dataset ever or it's just the model files I need to disclose?",
    "135467": "Hi, @installtt , no one will judge you.  Currently Leaderboard(LB) score is calculated on the first LB dataset, but then eventually final score will be evaluated on the other private dataset. This is why overfitting is always upon you. As I know this is just fun competition, so one will ask you for the model, but if you win payable competition, you  will need to disclose your model."
  },
  "source": "meta"
}