#!/usr/bin/env pythonfrom svm import *# the XOR problemlabels = [0, 1, 1, 0]samples = [[0, 0], [0, 1], [1, 0], [1, 1]]problem = svm_problem(labels, samples);size = len(samples)kernels = [LINEAR, POLY, RBF, SIGMOID]kname = ['linear','polynomial','rbf','sigmoid']param = svm_parameter(C = 10)for k in kernels:	param.kernel_type = k;	model = svm_model(problem,param)	errors = 0	for i in range(size):		prediction = model.predict(samples[i])		if (labels[i] != prediction):			errors = errors + 1	print "##########################################"	print " kernel %s: error rate = %d / %d" % (kname[param.kernel_type], errors, size)	print "##########################################"