Pattern Recognition I Unsupervised Analysis

It is the last statement in the challenge facing us that distinguishes the 

Meuzelaar and T.L. Isenhour, Computer Enhanced Analytical Spectroscopy , Plenum Press, New York, USA, 1987. 

Although the human brain is excellent at recognizing and classifying patterns and shapes, it performs less well if an object is represented by a numerical list of attributes, and much analytical data is acquired and presented in such a form. Consider the data shown in Table 4.1, obtained from an analysis of a series of alloys. This is only a relatively small data set but it may not be immediately apparent that these samples can be organized into well-defined groups defining the type or class of alloy according to their composition. The data from Table 

It is the last statement in the challenge facing us that distinguishes the techniques studied here from supervised pattern recognition schemes to be examined in Chapter 5. In supervised pattern recognition, a training set is identified with which the parent class or group of each sample is known, and this information is used to develop a suitable discriminant function with which 

---

Pattern Recognition I Unsupervised Analysis For the test data,... 

---