| The ease with which we recognize a face, understand spoken words, read handwritten characters, identify car keys in our pocket by feel and decide whether an apple is ripe by its smell belies the astoundingly complex processes that underlie these acts of pattern recognition. Pattern recognition—the art of processing raw data for categorization—is crucial to survival. And biology has produced this ability through evolution over tens of millions of years. It is natural that we should seek to design and build machines that can recognize patterns. From automated speech recognition, fingerprint identification, optical character recognition, DNA sequence identification and much more, it is clear that reliable accurate pattern recognition by machine would be immensely useful.; We have investigated two areas of machine-based pattern recognition, following the Artificial Neural Network approach. One is from the life sciences and the other from engineering. In life sciences, we considered the problem of cardiac auscultation from pediatric cardiology. This is a significant problem in pediatric cardiology, because of the high rate of incidence of heart murmurs in the pediatric population (reportedly 77% to 95%), of which only a small fraction arise from congenital heart disease. Using a variety of novel enhancements, we were able to develop a classifier with unprecedented accuracy to distinguish between innocent and pathological heart murmurs. These enhancements are: the smart-threshold rule, the committee of experts' scheme and the knowledge-access optimization scheme.; In engineering, we considered an example from manufacturing process control. Working with a mock reactor for the chemical vapor deposition process from semiconductor manufacturing, we developed an ANN classifier to identify process disruptions from the spatial patterns in deposition thickness variation. It was shown that a classifier could discriminate between process events with 100% accuracy. To achieve this high accuracy, the decision threshold rule and the committee of experts' scheme were applied. |
