Biometric Authentication: A Machine Learning Approach

3.1. Introduction

Learning networks are commonly categorized in terms of supervised and unsupervised networks. In unsupervised learning, the training set consists of input training patterns only. In contrast, in supervised learning networks, the training data consist of many pairs of input/output patterns. Therefore, the learning process can benefit greatly from the teacher's assistance. In fact, the amount of adjustment of the updating coefficients often depends on the difference between the desired teacher value and the actual response. As demonstrated in Chapter 5, many supervised learning models have been found to be promising for biometric authentication; their implementation often hinges on an effective data-clustering scheme, which is ...