Integration of Statistical and Neural Approaches
The sheer number of statistical and neural network-based classification methods induces the important and practical question: “Which classification method or methods are superior?” In this chapter we address the similarities and differences between statistical pattern classification methods and nonparametric (structural) neural network classification methods. Rather than focusing on the confrontation between these two classification paradigms, we focus on integrating them in order to utilise their particular strengths and properties and to diminish their shortcomings.
KeywordsStatistical Classifier Classification Rule Generalisation Error Distribution Density Function Neural Classifier
Unable to display preview. Download preview PDF.