Warning: this project is in very early stages and may change a lot.
DL4J wrapper for WEKA. Original code written by Mark Hall. This package currently introduces a new classifier,
Dl4jMlpClassifier, which allows arbitrary-depth MLPs to be built with a degree of flexibility (e.g. type of weight initialisation,
loss function, gradient descent algorithm, etc.).
Not many tests have been written for this classifier yet, so expect it to be quite buggy!
Simply run the
build.sh script in the core directory. This assumes:
- You have Ant and Maven installed.
weka.jarfile resides somewhere in your Java classpath. The latest and greatest WEKA installation is highly recommended; you can get the .jar of the nightly snapshot here.
An example script is provided that can be run on the Iris dataset in the
java -Xmx5g -cp $WEKA_HOME/weka.jar weka.Run \ .Dl4jMlpClassifier \ -S 0 \ -layer "weka.dl4j.layers.DenseLayer -units 10 -activation tanh -init XAVIER" \ -layer "weka.dl4j.layers.OutputLayer -units 3 -activation softmax -init XAVIER -loss MCXENT" \ -iters 100 \ -optim STOCHASTIC_GRADIENT_DESCENT \ -updater NESTEROVS \ -lr 0.1 \ -momentum 0.9 \ -bs 1 \ -t ../datasets/iris.arff \ -no-cv
This trains a one-hidden-layer MLP with 10 units on the Iris dataset. Nesterov momentum is used in conjunction with SGD and the initial learning rate and momentum is set to 0.1 and 0.9, respectively. The network is trained for 100 iterations.
DL4J is not primarily intended for research purposes -- rather more commercial and convention endeavours -- and so for more research-oriented tasks, a library such as Theano should be used in conjunction with the WekaPyScript package which allows WEKA classifiers to be prototyped in Python.