{
  "id": 21150,
  "title": "Is there any golden rule for choosing kernel size and max pooling layer size... ",
  "url": "/competitions/state-farm-distracted-driver-detection/discussion/21150",
  "author_name": "",
  "post_date": "2016-05-23T09:44:59.973Z",
  "votes": null,
  "comment_count": 2,
  "views": 931,
  "content": "<p>Is there any golden rule for choosing kernel size and max pooling layer size...  Also how does image size, image colors etc affect loss... And what is batch_size and how it affects loss....</p>",
  "messages": [
    {
      "id": "121054",
      "postDate": "05/23/2016 09:44:59",
      "content": "<p>Is there any golden rule for choosing kernel size and max pooling layer size...  Also how does image size, image colors etc affect loss... And what is batch_size and how it affects loss....</p>",
      "rawMarkdown": "Is there any golden rule for choosing kernel size and max pooling layer size...  Also how does image size, image colors etc affect loss... And what is batch_size and how it affects loss....",
      "votes": null
    },
    {
      "id": "121104",
      "postDate": "05/23/2016 23:21:41",
      "content": "<p><a href=\"https://arxiv.org/pdf/1512.00567.pdf\">This paper</a> has some good discussion on network architecture and kernel choices.  I find it easier to start with a network architecture that has been demonstrated by someone else to learn good predictions for one of the standard datasets .</p>",
      "rawMarkdown": "[This paper][1] has some good discussion on network architecture and kernel choices.  I find it easier to start with a network architecture that has been demonstrated by someone else to learn good predictions for one of the standard datasets .\r\n\r\n\r\n  [1]: https://arxiv.org/pdf/1512.00567.pdf",
      "votes": null
    },
    {
      "id": "121150",
      "postDate": "05/24/2016 14:25:59",
      "content": "<p>Hi Shahnawaz,</p>\n\n<p>How about trying some famous architectures and brushing up them ?<br>\nOne of the most famous architectures is <a href=\"https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks.pdf\">AlexNet</a>.<br></p>",
      "rawMarkdown": "Hi Shahnawaz,\r\n\r\nHow about trying some famous architectures and brushing up them ?<br>\r\nOne of the most famous architectures is [AlexNet][1].<br>\r\n\r\n\r\n\r\n  [1]: https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks.pdf",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 121104,
      "author_name": "mikowals",
      "author_url": "",
      "post_date": "05/23/2016 23:21:41",
      "content": "<p><a href=\"https://arxiv.org/pdf/1512.00567.pdf\">This paper</a> has some good discussion on network architecture and kernel choices.  I find it easier to start with a network architecture that has been demonstrated by someone else to learn good predictions for one of the standard datasets .</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 121150,
      "author_name": "toshik",
      "author_url": "",
      "post_date": "05/24/2016 14:25:59",
      "content": "<p>Hi Shahnawaz,</p>\n\n<p>How about trying some famous architectures and brushing up them ?<br>\nOne of the most famous architectures is <a href=\"https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks.pdf\">AlexNet</a>.<br></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "121054": "Is there any golden rule for choosing kernel size and max pooling layer size...  Also how does image size, image colors etc affect loss... And what is batch_size and how it affects loss....",
    "121104": "[This paper][1] has some good discussion on network architecture and kernel choices.  I find it easier to start with a network architecture that has been demonstrated by someone else to learn good predictions for one of the standard datasets .\r\n\r\n\r\n  [1]: https://arxiv.org/pdf/1512.00567.pdf",
    "121150": "Hi Shahnawaz,\r\n\r\nHow about trying some famous architectures and brushing up them ?<br>\r\nOne of the most famous architectures is [AlexNet][1].<br>\r\n\r\n\r\n\r\n  [1]: https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks.pdf"
  },
  "source": "meta"
}