A pattern is a finite string of constant and variable symbols. The erasing language generated by a pattern p is the set of all strings that can be obtained by substituting (possib...
In this paper, we present a kernel trick embedded Gaussian Mixture Model (GMM), called kernel GMM. The basic idea is to embed kernel trick into EM algorithm and deduce a parameter ...
The subject of this paper are algorithms for measuring the similarity of patterns of line segments in the plane, a standard problem in, e.g., computer vision, geographic informati...
Solomonoff unified Occam’s razor and Epicurus’ principle of multiple explanations to one elegant, formal, universal theory of inductive inference, which initiated the field...
Abstract. A crucial problem in machine learning is to choose an appropriate representation of data, in a way that emphasizes the relations we are interested in. In many cases this ...