I have this huge dimensional data with 31200 features and only 6000 examples. I want to learn a neural network that can find the non linear relation between the input and outputs. However, I have huge number of features. So wouldn't it cause problems in neural network?


You can try Dimensionality Reduction Techniques like PCA or LDA.

There's this survey by Fodor, but it involves some math, so you can look into this tutorial as well.

Another thing you can do is Feature Selection in the preprocessing step. You can look into this paper by Nguyen.

Also if you want an easy way out, you can use the Software Package from ASU. They have a fairly well documented report on it as well.

Hope this somewhat helps. Have fun Learning :)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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