TRAINING PATTERN REPLICATION AND WEIGHTED CLASS ALLOCATION IN ARTIFICIAL NEURAL-NETWORK CLASSIFICATION

In some image classifications the importance of classes varies, and it is desirable to weight allocation to selected classes. Often the desi re is to weight allocation in favour of classes that are abundant in t he area represented by an image at the expense of the less abundant cl asses. If there is prior knowledge on the distribution of class occurr ence, this weighting can be achieved with widely used statistical clas sifiers by setting appropriate a priori probabilities of class members hip. With an artificial neural network, the incorporation of prior kno wledge is more problematic. An approach to weight class allocation in an artificial neural network classification by replicating selected tr aining patterns is presented. This investigation focuses on a series o f classifications in which some classes were more abundant than others , but the same number of training cases were available for each class. By replicating the training patterns of abundant classes the represen tation of the abundant classes in the training set is increased, refle cting more closely the relative abundance of the classes in an image. Significant increases in classification accuracy were obtained by repl icating the training patterns of abundant classes. Furthermore, in com parison against a discriminant analysis for the classification of synt hetic aperture radar imagery, the results showed that training pattern replication could be used to weight class allocation with an effect s imilar to that of incorporating a priori probabilities of class member ship into the discriminant analysis, and resulted in a significant 20. 88%, increase in classification accuracy. This increase in classificat ion accuracy was obtained without any new information, but was the res ult of making fuller use of what was available.