模式识别 第三章 模式分类.pdf

Chapter 3 Distance Measures for Pattern Classification 1 Distance Based Classification  Distance based classification is the most common type of pattern recognition technique  Concepts are a basis for other classification techniques 2 Distance Based Classification  First we will look at choosing a class prototype A prototype is a sample or pattern which represents the class  Then we will look at how to calculate the distance from a new pattern that we are trying to classify to the class using the prototype 3 Outline  Euclidean Distance Classifiers  Prototype Selection  Distance Metrics  Orthonormal Whitening  Minimum Intra-Class Distance (MICD) Metric  Properties of the Metric  MICD Classifier, Decision Boundaries, and Regions 4 §3.1 Euclidean Distance Classifiers  The decision rule based on this metric is called the Minimum Euclidean Distance (MED) Classifier: 5 §3.1 Euclidean Distance Classifiers  The function is referred to as a discriminate function or decision function decision boundary 6 §3.1 Euclidean Distance Classifiers  For the MED classifier: This is the equation of a hyperplane with normal vector a distance from the origin 7 §3.1 Euclidean Distance Classifiers