{ "id": 552326, "title": "Discussion - Purpose of Autoencoder", "url": "/competitions/child-mind-institute-problematic-internet-use/discussion/552326", "author_name": "", "post_date": "2024-12-19T04:56:41.171922700Z", "votes": 2, "comment_count": 1, "views": 0, "content": "

I'm having some trouble interpreting the notebooks using auto encoders. Is there a preferred method of use for this data? Does it make sense to extract features of class A and use reconstruction error threshold to classify the unlabeled samples as A or B? Or just use it as a dimensionality reduction technique for the samples that are already labeled?

", "messages": [ { "id": "3075649", "postDate": "12/19/2024 04:56:41", "content": "

I'm having some trouble interpreting the notebooks using auto encoders. Is there a preferred method of use for this data? Does it make sense to extract features of class A and use reconstruction error threshold to classify the unlabeled samples as A or B? Or just use it as a dimensionality reduction technique for the samples that are already labeled?

", "rawMarkdown": "I'm having some trouble interpreting the notebooks using auto encoders. Is there a preferred method of use for this data? Does it make sense to extract features of class A and use reconstruction error threshold to classify the unlabeled samples as A or B? Or just use it as a dimensionality reduction technique for the samples that are already labeled?", "votes": null }, { "id": "3075695", "postDate": "12/19/2024 06:29:30", "content": "

Autoencoder is a vast topic and can't be explained in a single post, so I am redirecting you to a few references @maxrivera

\n
    \n
  1. https://www.geeksforgeeks.org/auto-encoders/
    2. \n
  2. https://www.datacamp.com/tutorial/introduction-to-autoencoders
    3. \n
  3. https://www.tensorflow.org/tutorials/generative/autoencoder
    4. \n
  4. https://www.ibm.com/think/topics/autoencoder
    5. \n
  5. https://www.mathworks.com/discovery/autoencoder.html
    6. \n
", "rawMarkdown": "Autoencoder is a vast topic and can't be explained in a single post, so I am redirecting you to a few references @maxrivera \n1. https://www.geeksforgeeks.org/auto-encoders/\n2. https://www.datacamp.com/tutorial/introduction-to-autoencoders\n3. https://www.tensorflow.org/tutorials/generative/autoencoder\n4. https://www.ibm.com/think/topics/autoencoder\n5. https://www.mathworks.com/discovery/autoencoder.html", "votes": null } ], "comments": [ { "id": 3075695, "author_name": "ravi20076", "author_url": "", "post_date": "12/19/2024 06:29:30", "content": "

Autoencoder is a vast topic and can't be explained in a single post, so I am redirecting you to a few references @maxrivera

\n
    \n
  1. https://www.geeksforgeeks.org/auto-encoders/
    2. \n
  2. https://www.datacamp.com/tutorial/introduction-to-autoencoders
    3. \n
  3. https://www.tensorflow.org/tutorials/generative/autoencoder
    4. \n
  4. https://www.ibm.com/think/topics/autoencoder
    5. \n
  5. https://www.mathworks.com/discovery/autoencoder.html
    6. \n
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "3075649": "I'm having some trouble interpreting the notebooks using auto encoders. Is there a preferred method of use for this data? Does it make sense to extract features of class A and use reconstruction error threshold to classify the unlabeled samples as A or B? Or just use it as a dimensionality reduction technique for the samples that are already labeled?", "3075695": "Autoencoder is a vast topic and can't be explained in a single post, so I am redirecting you to a few references @maxrivera \n1. https://www.geeksforgeeks.org/auto-encoders/\n2. https://www.datacamp.com/tutorial/introduction-to-autoencoders\n3. https://www.tensorflow.org/tutorials/generative/autoencoder\n4. https://www.ibm.com/think/topics/autoencoder\n5. https://www.mathworks.com/discovery/autoencoder.html" }, "source": "meta" }