0
$\begingroup$

I have a feature matrix with 1200 rows and 18930 columns. The matrix is sparse and the original paper has used a stacked denoising autoencoder for dimensionality reduction. Since I want to enhance the results, which newly published dimensionality reduction technique would be suiatable for my case?

$\endgroup$
2
  • $\begingroup$ What kind of data is it? Can you hyperlink the paper in your question? $\endgroup$
    – nwaldo
    Commented May 17 at 18:13
  • $\begingroup$ @nwaldo The values in the matrix represent term frequency of some genes name in a text hence they are continuous. $\endgroup$
    – Satarnejad
    Commented May 19 at 14:59

1 Answer 1

0
$\begingroup$

If your data is discrete, then you could try deep matrix factorization(DeepMF) which is most commonly used on collaborative filtering dataset(recommender systems). https://arxiv.org/abs/2010.00380

$\endgroup$
1
  • $\begingroup$ Is this approach applicable to continuous data, too? $\endgroup$
    – Satarnejad
    Commented May 19 at 15:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.