public class LinWengPlatt extends Object
This class implements an algorithm to fit a sigmoid function to the output of an SVM classifier. The algorithm is the one introduced by Hsuan-Tien Lin, Chih-Jen Lin, and Ruby C. Weng (who were in turn extending work by J. Platt), and this implementation is a direct translation of their pseudo-code as presented in Lin, Lin, Weng. A note on Platt's probabilistic outputs for support vector machines. In Machine Learning, vol. 68, pp. 267-276, 2007.
| Modifier and Type | Class and Description |
|---|---|
static class |
LinWengPlatt.ConvergenceFailure |
| Constructor and Description |
|---|
LinWengPlatt() |
public LinWengPlatt()
public static Sigmoid fit(double[] decisionValues, boolean[] labels) throws LinWengPlatt.ConvergenceFailure
LinWengPlatt.ConvergenceFailureCopyright © 2014. All rights reserved.