A Malsburg learning back propagation combination for handwritten alpha numeral recognition

摘要

A combination of Malsburg Learning BP Network for handwritten alpha numeral identification has been designed and developed. The network combination has been used to train a set of standard data to recognize handwritten alpha numerals. With Holdout Method a separate labeled test data set has been used to measure the performance of the system in terms of accuracy, precision, recall and finally the f-score. The performance of the system is appreciable. The total time required for learning and performance evaluation is appreciably small, also the time taken to identify individual alpha numerals is small. Thus the present handwritten alpha numerals identification system is efficient, effective and fast.