Mixture of discriminative learning experts of constant sensitivity for automated cytology screening

One practical objective in an automated cytology screening task is to obtain as high as possible specificity (the percentage of normal slides being classified as normal) while attaining acceptable (predefined) constant sensitivity. In this paper, we propose a new learning algorithm which continuously improves the specificity while maintaining constant sensitivity for pattern classification problems. We further propose to integrate the pre-trained networks with constant sensitivities into the mixture of experts (MOE) network configuration. This enables each trained expert to be responsive to specific subregions of the input spaces with minimum ambiguity and thus produces better performance