In this paper, we propose a new context-sensitive Bayesian learning algorithm. By modeling the distributions of data locations by a mixture of Gaussians, the new algorithm can uti...
Abstract— This article presents an efficient and mature visionbased navigation algorithm based on a sensory-motor learning. Neither Cartesian nor topological map is required, bu...
Christophe Giovannangeli, Philippe Gaussier, Gael ...
For almost a century, the Harvard Business School has used case studies as the basis for experiential learning in both MBA and Executive Education courses. This article presents a...
The recently introduced online confidence-weighted (CW) learning algorithm for binary classification performs well on many binary NLP tasks. However, for multi-class problems CW l...
: High capacity associative neural networks can be built from networks of perceptrons, trained using simple perceptron training. Such networks perform much better than those traine...