Evaluation and comparison of pattern classifiers for chemical applications: adaptive digital learning networks and linear discriminants

Abstract

We have refrained from value judgments such as "good" or "better than" in discussing both the classifiers and the evaluators. The actual characteristics which make a classifier suitable for a particular application may depend on the application itself, as, for example, when the penalty associated with misclassification of a class member is not the same as for misclassification of a nonmember.24 Although the percent correct prediction should probably be abandoned as a measure of the performance of binary classifiers. the other measures discussed can be useful in developing a total picture of relative and absolute classifier performance.Abstract: Recent research on the use of adaptive networks of digital learning elements for chemical pattern recognition has stressed the high performance of such classifiers and their applicability to linearly inseparable data. In the present work, we apply a new performance measure, the figure of merit, and a large set of test data in a rigorous evaluation of the performance of digital learning networks. The results herein reported show that, when confronted with a large data set selected without particular consideration of the peculiarities of the network, the digital learning network continues to give good performance, although this performance is substantially below the levels previously reported. A comparison of the performance of the digital learning network classifiers with that of a set of linear discriminant functions indicates similar levels of performance for the two types of classifier.

Data Evaluation in High-Performance Liquid Chromatography-Diode-Array Detection-Fluorescence Detection by Information Theory

“…Evaluation of analytical data is usually focused on the precision and accuracy of determinations during an analytical process. Depending on their sizes, and a criterion chosen to satisfy a standard quality, , these statistical parameters can be used in optimization of experimental parameters to yield maximum information from the process. Thus, the information theory is deeply involved in analytical chemistry, mainly as a consequence between an analytical and a communication process .…”

Estimation of the Information Content in Ultraviolet−Visible Molecular Absorption Spectrometry

“…The foundations of the information theory are due to Shannon . The first applications of probability theory and, later, of information theory in analytical chemistry are due to Kaiser and were related to the spectral analytical methods, , and after that, Soltzberg and co-workers used this theory to evaluate the pattern classifiers for chemical pattern recognition. , The application of this theory in analytical chemistry is mainly a consequence of the similarity between an analytical process and a data transmission process from a source to a receptor . Eckschlager's works are among the most useful contributions in this domain…”

Interactive Pattern Recognition in the Chemical Analysis Laboratory

“…t Presented at the Joint U.S.-Japan Seminar on "Computer-Assisted Chemical Research Design", Aug[16][17][18][19][20] 1976, Washington, D.C., sponsored by the National Science Foundation and the Japan Society for the Promotion of Science.242 Journal of Chemical Information and Computer Sciences, Vol. 17,…”