{
  "id": 68365,
  "title": "RSME for ANN? ",
  "url": "/competitions/ga-customer-revenue-prediction/discussion/68365",
  "author_name": "",
  "post_date": "2018-10-12T00:06:37.341212900Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hello, is there a way for us to get RSME for ANN ? </p>",
  "messages": [
    {
      "id": "402580",
      "postDate": "10/12/2018 00:06:37",
      "content": "<p>Hello, is there a way for us to get RSME for ANN ? </p>",
      "rawMarkdown": "Hello, is there a way for us to get RSME for ANN ?",
      "votes": null
    },
    {
      "id": "402626",
      "postDate": "10/12/2018 02:47:26",
      "content": "<p>Using keras (backend TensorFlow) it will look like this: </p>\n\n<pre><code>import keras.backend as K\ndef rmse(y_true, y_pred):\n        return K.sqrt(K.mean(K.square(y_pred - y_true), axis=-1)) \n</code></pre>\n\n<p>Hope this helps. </p>",
      "rawMarkdown": "Using keras (backend TensorFlow) it will look like this: \n\n    import keras.backend as K\n    def rmse(y_true, y_pred):\n            return K.sqrt(K.mean(K.square(y_pred - y_true), axis=-1)) \n\nHope this helps.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 402626,
      "author_name": "leonidg",
      "author_url": "",
      "post_date": "10/12/2018 02:47:26",
      "content": "<p>Using keras (backend TensorFlow) it will look like this: </p>\n\n<pre><code>import keras.backend as K\ndef rmse(y_true, y_pred):\n        return K.sqrt(K.mean(K.square(y_pred - y_true), axis=-1)) \n</code></pre>\n\n<p>Hope this helps. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "402580": "Hello, is there a way for us to get RSME for ANN ?",
    "402626": "Using keras (backend TensorFlow) it will look like this: \n\n    import keras.backend as K\n    def rmse(y_true, y_pred):\n            return K.sqrt(K.mean(K.square(y_pred - y_true), axis=-1)) \n\nHope this helps."
  },
  "source": "meta"
}