We present a novel clustering method using HMM parameter space and eigenvector decomposition. Unlike the existing methods, our algorithm can cluster both constant and variable leng...
In nearest neighbor searching we are given a set of n data points in real d-dimensional space, d , and the problem is to preprocess these points into a data structure, so that give...
Abstract. This paper studies the properties and performance of models for estimating local probability distributions which are used as components of larger probabilistic systems â€...
Kristina Toutanova, Mark Mitchell, Christopher D. ...
CFT is a new constraint system providing records as logical data structure for constraint (logic) programming. It can be seen as a generalization of the rational tree system emplo...
Given a set D = {d1, d2, ..., dD} of D strings of total length n, our task is to report the "most relevant" strings for a given query pattern P. This involves somewhat mo...