Extensions of transductive learning for distributed ensemble classification and application to biometric authentication

Abstract

We consider ensemble classification when there is no common labeled data for designing the function which aggregates classifier decisions. In recent work, we dubbed this problem distributed ensemble classification, addressing when local classifiers are trained on different (e.g., proprietary, legacy) databases or operate on different sensing modalities. Typically, fixed ( untrained) rules of classifier combination such as voting methods are used in this case. However, these may perform poorly, especially when (i) the local class priors, used in training, differ from the true (test batch) priors and (ii) classifier decisions are statistically dependent. Alternatively, we proposed several transductive methods, optimizing the combining rule for objective functions measured on the test batch. We proposed both maximum likelihood (ML) and constraint-based (CB) objectives and found that CB achieved superior performance. Here, we develop CB extensions (i) for sequential decisionmaking and (ii) exploiting additional class information contained in the local classifier feature vectors. The new sequential method is applied to biometric authentication. We demonstrate these new CB methods achieve better ensemble decision accuracy than methods which apply fixed rules in combining classifier decisions.