Global likelihood maximization is an important aspect of many statistical analyses. Often the likelihood function is highly multi-extremal. This presents a significant challenge t...
This paper presents a minimum classification error (MCE) training approach for improving the accuracy of multi-class support vector machine (SVM) classifiers. We have applied th...