The user can choose from the following available layers:Ī layer can be further configured, e.g. The layer specification option lets the user specify the sequence of layers that build the neural network architecture: The network configuration option exposes further hyperparameter tuning:
The Dl4jMlpClassifier can be configured as shown below: Make sure your WEKA_HOME environment variable is set. Simple examples are given in the examples section for the Iris dataset and the MNIST dataset. The main classifier exposed by this package is named Dl4jMlpClassifier. If you are new to Weka, a good resource to get started is the Weka manual.Īs most of Weka, the WekaDeeplearning4j's functionality is accessible in three ways:Īll three ways are explained in the following.
0 kommentar(er)
