Local Models
摘要
In the previous chapter, we have looked at several discriminant function based classification algorithms. They are parameterized, i.e., assuming data distributions for fitting, or having specific (linear) models to optimize (like in LDA and SVM).