Hidden Markov Models (HMMs) are statistical models of sequential data that have been used successfully in many machine learning applications, especially for speech recognition. Furthermore, in the last few years, many new and promising probabilistic models related to HMMs have been proposed. We first summarize the basics of HMMs, and then review several recent related learning algorithms and extensions of HMMs, including in particular hybrids of HMMs with artificial neural networks, Input-Output HMMs (which are conditional HMMs using neural networks to compute probabilities), weighted transducers, variable-length Markov models and Markov switching state-space models. Finally, we discuss some of the challenges of future research in this very active area. 1 Introduction Hidden Markov Models (HMMs) are statistical models of sequential data that have been used successfully in many applications in artificial intelligence, pattern recognition, speech recognition, and modeling of biological ...

Probabilistic networks, which provide compact descriptions of complex stochastic relationships among several random variables, are rapidly becoming the tool of choice for uncertain reasoning in artificial intelligence. We show that networks with fixed structure containing hidden variables can be learned automatically from data using a gradient-descent mechanism similar to that used in neural networks. We also extend the method to networks with intensionally represented distributions, including networks with continuous variables and dynamic probabilistic networks. Because probabilistic networks provide explicit representations of causal structure, human experts can easily contribute prior knowledge to the training process, thereby significantly improving the learning rate. Adaptive probabilistic networks (APNs) may soon compete directly with neural networks as models in computational neuroscience as well as in industrial and financial applications. 1 Introduction Intelligent systems, ...

A new algorithm for training Restricted Boltzmann Machines is introduced. The algorithm, named Persistent Contrastive Divergence, is different from the standard Contrastive Divergence algorithms in that it aims to draw samples from almost exactly the model distribution. It is compared to some standard Contrastive Divergence and Pseudo-Likelihood algorithms on the tasks of modeling and classifying various types of data. The Persistent Contrastive Divergence algorithm outperforms the other algorithms, and is equally fast and simple.

We present a framework for the rational analysis of elemental causal induction – learning about the existence of a relationship between a single cause and effect – based upon causal graphical models. This framework makes precise the distinction between causal structure and causal strength: the difference between asking whether a causal relationship exists and asking how strong that causal relationship might be. We show that two leading rational models of elemental causal induction, ∆P and causal power, both estimate causal strength, and introduce a new rational model, causal support, that assesses causal structure. Causal support predicts several key phenomena of causal induction that cannot be accounted for by other rational models, which we explore through a series of experiments. These phenomena include the complex interaction between ∆P and the base-rate probability of the effect in the absence of the cause, sample size effects, inferences from incomplete contingency tables, and causal learning from rates. Causal support also provides a better account of a number of existing datasets than either ∆P or causal power.

Distributed representations are attractive for a number of reasons. They offer the possibility of representing concepts in a continuous space, they degrade gracefully with noise, and they can be processed in a parallel network of simple processing elements. However, the problem of representing nested structure in distributed representations has been for some time a prominent concern of both proponents and critics of connectionism [Fodor and Pylyshyn 1988; Smolensky 1990; Hinton 1990]. The lack of connectionist representations for complex structure has held back progress in tackling higher-level cognitive tasks such as language understanding and reasoning. In this thesis I review connectionist representations and propose a method for the distributed representation of nested structure, which I call "Holographic Reduced Representations " (HRRs). HRRs provide an implementation of Hinton's [1990] "reduced descriptions". HRRs use circular convolution to associate atomic items, which are rep...

ABayesian network is a probabilistic representation for uncertain relationships, which has proven to be useful for modeling real-world problems. When there are many potential causes of a given e ect, however, both probability assessment and inference using a Bayesian network can be di cult. In this paper, we describe causal independence, a collection of conditional independence assertions and functional relationships that are often appropriate to apply to the representation of the uncertain interactions between causes and e ect. We show how the use of causal independence in a Bayesian network can greatly simplify probability assessment aswell as probabilistic inference. 1

We provide an introduction to the theory and use of variational methods for inference and estimation in the context of graphical models. Variational methods become useful as ecient approximate methods when the structure of the graph model no longer admits feasible exact probabilistic calculations. The emphasis of this tutorial is on illustrating how inference and estimation problems can be transformed into variational form along with describing the resulting approximation algorithms and their properties insofar as these are currently known. 1 Introduction The term variational methods refers to a large collection of optimization techniques. The classical context for these methods involves nding the extremum of an integral depending on an unknown function and its derivatives. This classical de nition, however, and the accompanying calculus of variation no longer adequately characterizes modern variational methods. Modern variational approaches have become indispensable tools in...

We introduce a method for learning Bayesian networks that handles the discretization of continuous variables as an integral part of the learning process. The main ingredient in this method is a new metric based on the Minimal Description Length principle for choosing the threshold values for the discretization while learning the Bayesian network structure. This score balances the complexity of the learned discretization and the learned network structure against how well they model the training data. This ensures that the discretization of each variable introduces just enough intervals to capture its interaction with adjacent variables in the network. We formally derive the new metric, study its main properties, and propose an iterative algorithm for learning a discretization policy. Finally, we illustrate its behavior in applications to supervised learning. 1 INTRODUCTION Bayesian networks provide efficient and effective representation of the joint probability distribution over a set ...

Many real world learning problems are best characterized by an interaction of multiple independent causes or factors. Discovering such causal structure from the data is the focus of this paper. Based on

This paper introduces a new statistical approach to partitioning text automatically into coherent segments. Our approach enlists both short-range and long-range language models to help it sniff out likely sites of topic changes in text. To aid its search, the system consults a set of simple lexical hints it has learned to associate with the presence of boundaries through inspection of a large corpus of annotated data. We also propose a new probabilistically motivated ' error metric for use by the natural language processing and information retrieval communities, intended to supersede precision and recall for appraising segmentation algorithms. Qualitative assessment of our algorithm as well as evaluation using this new metric demonstrate the effective- ness of our approach in two very different domains, Wall Street Journal articles and the TDT Corpus, a collection of newswire articles and broadcast news transcripts.