Abstract-- Feature selection is an important method for improving the efficiency and accuracy of text categorization algorithms by removing redundant and irrelevant terms from the ...
This paper presents a general, trainable system for object detection in unconstrained, cluttered scenes. The system derives much of its power from a representation that describes a...
This paper presents the first comprehensive exploration of the largest Chinese online knowledge sharing communityBaidu Knows. With analyzing 5.2 millions questions and 2.7 million...
The ability to represent knowledge about space and its position therein is crucial for a mobile robot. To this end, topological and semantic descriptions are gaining popularity fo...
This paper identifies five distinct mechanisms by which a population-based algorithm might have an advantage over a solo-search algorithm in classical optimization. These mechanism...