{
  "id": 61300,
  "title": "Weighted loss function in keras ",
  "url": "/competitions/data-science-bowl-2018/discussion/61300",
  "author_name": "Amirreza Mahbod",
  "post_date": "2018-07-17T13:04:13.093000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi,\nI wonder to know if there is any loss function implementation in Keras as was used in the original U-net paper (weighted cross entropy loss function to add extra penalty in the touching borders)</p>\n\n<p>Thanks in advance</p>",
  "messages": [
    {
      "id": 358062,
      "postDate": "2018-07-17T13:04:13.093Z",
      "content": "<p>Hi,\nI wonder to know if there is any loss function implementation in Keras as was used in the original U-net paper (weighted cross entropy loss function to add extra penalty in the touching borders)</p>\n\n<p>Thanks in advance</p>",
      "rawMarkdown": "Hi,\nI wonder to know if there is any loss function implementation in Keras as was used in the original U-net paper (weighted cross entropy loss function to add extra penalty in the touching borders)\n\nThanks in advance",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "358062": "Hi,\nI wonder to know if there is any loss function implementation in Keras as was used in the original U-net paper (weighted cross entropy loss function to add extra penalty in the touching borders)\n\nThanks in advance"
  }
}