Discriminant Function Revisited for Incremental Learning

Discriminant function is commonly and effective methodology for solving classification problems. However, it is computationally efficient when all features are considered simultaneously. But sometimes all the features do not contribute significantly to classification. Also the noisy attributes sometimes may decrease the accuracy of classifier. So before classification feature selection is used as a pre-processing step. When the features are added one by one in forward feature selection method using batch mode, to compute discriminant function involves huge computation. In this paper, an incremental discriminant function for multivariate normal distribution datasets is proposed. The proposed incremental discriminant function is computationally efficient over batch discriminant function in terms of time. The effectiveness of the proposed incremental discriminant function has been demonstrated through experiments on different datasets. It is found on the basis of experiments that the incremental discriminant function has an equivalent power compared to batch discriminant function in terms of classification accuracy. However, the proposed incremental discriminant function has very high speed efficiency in comparison to batch discriminant function.