Software

Support Vector Machines (SVMs) are a new generation of supervised learning Algorithms, which can be applied succesfully to classify texts (e.g. spam filters), in character recognition, to classifiy signals and in many other areas of application.

To demonstrate the operating principle of SVMs, we implemented a simple demonstration program in Python.

This program allows you to generate two dimensional classification problems and to train a SVM with these data. The software draws the computed decision line together with the identified support vectors.

See

www.procoders.net/en/Procoders/open_source/pcSVMdemo