Deep learning in H2O

I did not find an easy-to-use native deep neural network package on R.
IMO darch is actually a good one. But it still takes some effort to pre-process data, tune the parameters, etc.

On the other hand, H2O platform is very easy to use. And it got a decent R integration package.

Below is a quick sample to model the MNIST data set using deep neural network (784 x 400 x 400 x 10, a small network). Here I uses the data downloaded from Kaggle. H2O platform need to be installed beforehand.

It would easily get 97% accuracy, and rank 300+ in Kaggle leader board.

Note the implementation in H2O is rather modern, compared to e.g. darch. It uses ReLU as activation function of hidden unit. It trains with dropout, which contributes to generalization. The implementation utilizes concurrency without locking.

For further knowledge, I recommend A ‘Brief’ History of Neural Nets and Deep Learning as an introduction material.

This entry was posted in Computer and Internet, Machine Learning, Science and tagged . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s