Package com.imsl.test.example.datamining.neural

Class MultiClassificationEx1

java.lang.Object

com.imsl.test.example.datamining.neural.MultiClassificationEx1

All Implemented Interfaces:

Serializable

public class MultiClassificationEx1
extends Object
implements Serializable

Trains a 3-layer network to Fisher's iris data.

This example trains a 3-layer network using Fisher's Iris data with four continuous input attributes and three output classifications. Fisher's Iris data is perhaps the best known data set in the pattern recognition literature. Published in 1936 Fisher's paper is a classic in the field. The data set contains 3 classes of 50 instances each, where each class refers to a type of iris plant. The continuous attributes are measurements of petal length, petal width, sepal length, sepal width.

The structure of the network consists of four input nodes and three layers, with four perceptrons in the first hidden layer, three perceptrons in the second hidden layer and three in the output layer.

The four input attributes represent

1. Sepal length
2. Sepal width
3. Petal length
4. Petal width

The output attribute represents the class of the iris plant and are encoded using binary encoding:

1. Iris Setosa
2. Iris Versicolour
3. Iris Virginica

With 4 inputs, 4 and 3 perceptrons in the hidden layers, and 3 perceptrons in the output layer there are a total of 47 weights in this network, including the bias weights. All hidden layers use the logistic activation function. Since the target output is multi-classification the softmax activation function is used in the output layer and the MultiClassification error function class is used by the trainer. The error class MultiClassification combines the cross-entropy error calculations and the softmax function.

See Also:

• Code
• Output
• Serialized Form

Constructor Summary

Constructors

Constructor	Description
MultiClassificationEx1()

Method Summary

Modifier and Type	Method	Description
static void	main(String[] args)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

MultiClassificationEx1

public MultiClassificationEx1()

Method Details

main

public static void main(String[] args)
                 throws Exception

Throws:

Exception