Hi, sorry if this is the wrong place to post. Very new to SVM, was wondering if there were any examples of a one-class classifier (ie. trained on normal data, tested on normal and novel data) using LibSVM in Java. Is the only thing we need to set to make it a one-class classifier "param.svm_type = svm_parameter.ONE_CLASS;"?
Also, are there any tools to make one-class classification selection of parameters easier? (Similar to easy.py for what I'm assuming is multi-class or binary classification). Thank you.
Please check README. If you write a java code to call libsvm, besides svm_type you also need to choose some parameters such as nu, kernel, kernel parameters etc.
Hi, sorry if this is the wrong place to post. Very new to SVM, was wondering if there were any examples of a one-class classifier (ie. trained on normal data, tested on normal and novel data) using LibSVM in Java. Is the only thing we need to set to make it a one-class classifier "param.svm_type = svm_parameter.ONE_CLASS;"?
Also, are there any tools to make one-class classification selection of parameters easier? (Similar to easy.py for what I'm assuming is multi-class or binary classification). Thank you.