We propose a discriminative learning approach for fusing multichannel sequential data with application to detect unsafe driving patterns from multi-channel driving recording data....
Real-world face recognition systems often have to face the single sample per person (SSPP) problem, that is, only a single training sample for each person is enrolled in the datab...
The process of diagnosis involves learning about the state of a system from various observations of symptoms or findings about the system. Sophisticated Bayesian (and other) algor...
We consider the general problem of learning from labeled and unlabeled data. Given a set of points, some of them are labeled, and the remaining points are unlabeled. The goal is to...
Some discourse structures such as enumerative structures have typographical, punctuational and laying out characteristics which (1) make them easily identifiable and (2) convey hi...