Deep Learning on the JVM

DL4J is a pretty awesome open source project that works with Spark and Hadoop.  Deep Learning 4J also works as a YARN app! It includes Text, NLP, Canova Vectorization Lib for ML, Scientific computing for the JVM, distributed with clusters, and works with CUDA GPU kernels.

Take a look at the Quick Start. 

DL4J is used for anomaly detection (fraud detection), recommender systems, predictive analytics with logs and image recognition.

In a related open source project, Skymind built a numerical computing library ND4J, or n-dimensional arrays for Java, essentially porting Numpy to the JVM. ND4J powers Deeplearning4j, since deep-learning relies heavily on linear algebra and calculus. And the real work of ND4J is done in C++ with libnd4j. . 

It is very easy to get started and run their code in IntelliJ, even on an average machine the results are quick.

DL4J Resources

• DL4J Github

• ND4J Github

• LibND4J Github

• GitterIM Chat with the DL4J Team

• Full Documentation

• DL4J Wiki Entry

• DeepLearning4J on Spark 

• NLP with DL4J

• Word2Vec in DL4J

• DL4J:  Full instructions for setup

• Scala Examples

• Introduction to Deep Neural Networks

• A Beginner’s Tutorial for Restricted Boltzmann Machines

Skymind and Hortonworks will be presenting an in-depth DL4j + Spark on HDP talk at the 2016 Hadoop Summit in San Jose, CA

Simple Setup to Run the Examples (Must Have JDK, Maven, and IntelliJ)

git clone https://github.com/deeplearning4j/dl4j-0.4-examples.git

cd dl4j-0.4-examples/

mvn clean install

brew install homebrew/science/deeplearning4j-cli