I am interested in the existing hardware implementations of machine learning methods (classifiers such as SVM, Bayesian networks, HMM, GMM). Unfortunately, I have found only prototyping proposals and concept-proving reports published in the Internet. It gives me no sense about solutions used in the industry now.

What is the state of the art for hardware implementation of machine learning algorithms? Any public information about industry-oriented solutions?

asked May 24 '13 at 10:55

Piotr%20Semenov's gravatar image

Piotr Semenov
16226


One Answer:

Are you looking for this? I mean your question was either too obvious or to vague. But, have a look at what the guys at Grok, Eratz and Wise.io are doing. The latter even showed that you can have fast machine learning using Raspberry Pi. It depends on your problem.

PS: Since your problem is ill-posed, I believe you will have better answers if you constraint task and budget.

answered May 24 '13 at 18:18

edersantana's gravatar image

edersantana
155259

Your answer
toggle preview

powered by OSQA

User submitted content is under Creative Commons: Attribution - Share Alike; Other things copyright (C) 2010, MetaOptimize LLC.