Skip to content
This repository has been archived by the owner on Jul 10, 2021. It is now read-only.

Slow learning for high-dimensional input #237

Open
hollma opened this issue Jul 5, 2017 · 0 comments
Open

Slow learning for high-dimensional input #237

hollma opened this issue Jul 5, 2017 · 0 comments

Comments

@hollma
Copy link

hollma commented Jul 5, 2017

Dear developers,

first of all, thank you for this fine peace of free software!

I have been using scikit-neuralnetwork 0.7 with scikit-learn 0.18.2, theano 0.7.0, and lasagne 0.1, when I noticed that learning halfspaces seems to be quite slow when the examples are high-dimensional vectors, e.g. dim >= 1000. These dimensions are quite normal when using tfidf vectors in text classification settings.

A minimal working example (with runtime stats):
https://gist.github.com/hollma/f0d98bc5e58a6db34725dbce9ecdf9d1

Processing 500 training examples (100-dimensional) and validating the success on another 500 test examples took almost 14 seconds (on an Intel i7-4790 CPU running at 3.6 GHz).

What would you recommend? Are high-dimensional input vectors the wrong use-case for scikit-neuralnetwork, i.e. I should use some other library instead?

I am looking forward to your answer.

Best regards,
Mario

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant