{
  "id": 70851,
  "title": "what is loss and metric in Keras model.compile?",
  "url": "/competitions/ga-customer-revenue-prediction/discussion/70851",
  "author_name": "",
  "post_date": "2018-11-07T21:48:19.880249100Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>For this regression problem, I use mse as loss and rmse as metric in model.compile. Is it correct?\nOr do I need to use rmse as loss and accuracy as metric? </p>",
  "messages": [
    {
      "id": "417164",
      "postDate": "11/07/2018 21:48:19",
      "content": "<p>For this regression problem, I use mse as loss and rmse as metric in model.compile. Is it correct?\nOr do I need to use rmse as loss and accuracy as metric? </p>",
      "rawMarkdown": "For this regression problem, I use mse as loss and rmse as metric in model.compile. Is it correct?\nOr do I need to use rmse as loss and accuracy as metric?",
      "votes": null
    },
    {
      "id": "417165",
      "postDate": "11/07/2018 21:50:31",
      "content": "<p>And as I said, I use rmse as metric now. If the rmse of validation dataset is bigger than rmse of training dataset, does it mean my model is overfitting?</p>",
      "rawMarkdown": "And as I said, I use rmse as metric now. If the rmse of validation dataset is bigger than rmse of training dataset, does it mean my model is overfitting?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 417165,
      "author_name": "rachel115",
      "author_url": "",
      "post_date": "11/07/2018 21:50:31",
      "content": "<p>And as I said, I use rmse as metric now. If the rmse of validation dataset is bigger than rmse of training dataset, does it mean my model is overfitting?</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "417164": "For this regression problem, I use mse as loss and rmse as metric in model.compile. Is it correct?\nOr do I need to use rmse as loss and accuracy as metric?",
    "417165": "And as I said, I use rmse as metric now. If the rmse of validation dataset is bigger than rmse of training dataset, does it mean my model is overfitting?"
  },
  "source": "meta"
}