The concept of a learning object (LO) has spread quickly without a very specific universal definition, and though born originally from the idea of object oriented design, with a ...
With the goal to generate more scalable algorithms with higher efficiency and fewer open parameters, reinforcement learning (RL) has recently moved towards combining classical tec...
This paper presents a general learning framework for non-rigid registration of MR brain images. Given a set of training MR brain images, three major types of information are partic...
In this paper, we present an overview of generalized expectation criteria (GE), a simple, robust, scalable method for semi-supervised training using weakly-labeled data. GE fits m...
In the paper we present a generalized discriminative multiple instance learning algorithm (GD-MIL) for multimedia semantic concept detection. It combines the capability of the MIL...