nd4j

Hi Jordi - I'm sure you're right. You have an amazing resume. Please help us make ND4J/ND4S better. https://github.com/deeplearning4j/nd4j Nasa JPL is using the framework for climatic modeling, so maybe we can help save the planet together. :) One response I would have to both your comment and @alephone's rewording, which is perfectly correct, is that they are incomprehensible to most people. So the question is: How do you bridge the gap to make non-mathematicians understand? The other would be: It's hard to talk about math. There's always slippage when you try to speak in an everyday way about technical things. I'm not sure how to get around that. Too much precision, in this case, undoes the goal of reaching another audience.

Disclaimer, it's my project, but I run an open source project called deeplearning4j, who's algorithms have a hardware abstraction layer built in to them called nd4j. You get numpy on the jvm and hardware as a jar file. Deeplearning4j itself is built on top of that. Would love to help spread deep learning to different runtimes. Here's the current work being done on opencl: https://github.com/deeplearning4j/nd4j/tree/master/nd4j-jocl... We'd love to get this finished. Bit more to do yet though...definitely looking for contributors here. You'll get opencl neural nets for free.

* Deeplearning4j - distributed DL for the JVM https://github.com/deeplearning4j/deeplearning4j * ND4J - Scientific computing for the JVM https://github.com/deeplearning4j/nd4j * Vectorization lib for Machine Learning Tools https://github.com/deeplearning4j/Canova