The information bottleneck is an information theoretic framework, extending the classical notion of minimal sufficient statistics, that finds concise representations for an ‘input’ random variable that are as relevant as possible for an ‘output’ variable. This framework has been used successfully in various supervised and unsupervised applications. However, its learning theoretic properties and...

This paper argues that Bayesian probability theory is a general method for machine learning. From two well-founded axioms, the theory is capable of accomplishing learning tasks that are incremental or non-incremental, supervised or unsupervised. It can learn from different types of data, regardless of whether they are noisy or perfect, independent facts or behaviors of an unknown machine. These...

Deteministic finit state (DFA) automata have emerged as an effective tools for agent modeling applications. The problem of automata learning is to determine a DFA from a series of observation and has recently been studied extensively and a number of algorithms has been proposed. These algorithms can be divided into groups : supervised and unsupervised . In supervised algorithms, we have access ...

Active contours, very popular in image segmentation, suffer from delicate adjustments of many parameters. We propose to carry out these adjustments using genetic algorithm. Here an active contour is implemented using a greedy algorithm. Within this framework, two approaches are presented. A supervised approach which delivers a global set of parameters. In this case the greedy algorithm is invol...

In this paper we propose an instance based method for lexical entailment and apply it to automatic ontology population from text. The approach is fully unsupervised and based on kernel methods. We demonstrate the effectiveness of our technique largely surpassing both the random and most frequent baselines and outperforming current state-of-the-art unsupervised approaches on a benchmark ontology...

The Semantic Web is an extension of the current web in which information is given well-defined meaning. The perspective of Semantic Web is to promote the quality and intelligence of the current web by changing its contents into machine understandable form. Therefore, semantic level information is one of the cornerstones of the Semantic Web. The process of adding semantic metadata to web resourc...

We have been investigating the use of kernel methods to improve conventional linear adaptation algorithms for fast adaptation, when there are less than 10s of adaptation speech. On clean speech, we had shown that our new kernel-based adaptation methods, namely, embedded kernel eigenvoice (eKEV) and kernel eigenspace-based MLLR (KEMLLR) outperformed their linear counterparts. In this paper, we s...

We derive a spectral method for unsupervised learning of Weighted Context Free Grammars. We frame WCFG induction as finding a Hankel matrix that has low rank and is linearly constrained to represent a function computed by inside-outside recursions. The proposed algorithm picks the grammar that agrees with a sample and is the simplest with respect to the nuclear norm of the Hankel matrix.

Low-rank matrix decomposition has gained great popularity recently in scaling up kernel methods to large amounts of data. However, some limitations could prevent them from working effectively in certain domains. For example, many existing approaches are intrinsically unsupervised, which does not incorporate side information (e.g., class labels) to produce task specific decompositions; also, the...