Sliding window classifiers are among the most successful and widely applied techniques for object localization. However, training is typically done in a way that is not specific to...
CoIL challenge 2000 was a supervised learning contest that attracted 43 entries. The authors of 29 entries later wrote explanations of their work. This paper discusses these repor...
We present a new algorithm for minimizing a convex loss-function subject to regularization. Our framework applies to numerous problems in machine learning and statistics; notably,...
Constructive Induction is the process of transforming the original representation of hard concepts with complex interaction into a representation that highlights regularities. Mos...
Abstract— Many games require opponent modelling for optimal performance. The implicit learning and adaptive nature of evolutionary computation techniques offer a natural way to d...