Anomaly detection is a key element of intrusiondetection and other detection systems in which perturbations of normal behavior suggest the presence of intentionally or unintentionally induced attacks, faults, defects, etc. Because most anomaly detectors are based on probabilistic algorithms that exploit the intrinsic structure, or regularity, embedded in data logs, a fundamental question is whether or not such structure influences detection performance. If detector performance is indeed a function of environmental regularity, it would be critical to match detectors to environmental characteristics. In intrusion-detection settings, however, this is not done, possibly because such characteristics are not easily ascertained. This paper introduces a metric for characterizing structure in data environments, and tests the hypothesis that intrinsic structure influences probabilistic detection. In a series of experiments, an anomaly-detection algorithm was applied to a benchmark suite of 165 c...

Abstract. We present a new algorithm for Bayesian network structure learning, called Max-Min Hill-Climbing (MMHC). The algorithm combines ideas from local learning, constraint-based, and search-and-score techniques in a principled and effective way. It first reconstructs the skeleton of a Bayesian network and then performs a Bayesian-scoring greedy hill-climbing search to orient the edges. In our extensive empirical evaluation MMHC outperforms on average and in terms of various metrics several prototypical and state-of-the-art algorithms, namely the PC, Sparse Candidate, Three Phase Dependency Analysis, Optimal Reinsertion, Greedy Equivalence Search, and Greedy Search. These are the first empirical results simultaneously comparing most of the major Bayesian network algorithms against each other. MMHC offers certain theoretical advantages, specifically over the Sparse Candidate algorithm, corroborated by our experiments. MMHC and detailed results of our study are publicly available at

Visual stimuli contain a limited amount of information that could potentially be used to perform a given visual task. At successive stages of visual processing, some of this information is lost and some is transmitted to higher stages. This article describes a new analysis, based on the concept of the ideal observer in signal detection theory, that allows one to trace the (low of discrimination information through the initial physiological stages of visual processing, for arbitrary spatio-chromatic stimuli. This ideal-observer analysis provides a rigorous means of measuring the information content of visual stimuli and of assessing the contribution of specific physiological mechanisms to discrimination performance. Here, the analysis is developed for the physiological mechanisms up to the level of the photoreceptor. It is shown that many psychophysical phenomena previously attributed to neural mechanisms may be explained by variations in the information content of the stimuli and by preneural mechanisms. The purpose of vision is to extract and represent information about the physical environment from the light that is emitted, transmitted, or reflected by objects and surfaces. In order to extract useful information, a visual system must be able to encode

This report investigates the "decision lanisio es" that characterize several forms of biometric decision makinn The issues discussed inIP/PF (i) Estimatin the degrees-of-freedom associated with different biometrics, as a way ofmeasurin theranFfl3#9N an complexity(an therefore the unWflWW#9Nfl of their templates. (ii) The conflflP#9NflY of combin/I more than on biometric test to arrive at a decision (iii) The requiremen ts for performin iden tification by large-scale exhaustive database search, as opposed to mere verification bycomparison againr a sin;I template. (iv)ScenWP3F for Biometric Key Cryptography (the use of biometrics forenPflW/#9N of messages). These issues are conFFYI#9 here in abstract form, but where appropriate, the particular example of iris recognflfl#9 is used asan illustration Aun;FflI# theme of all four sets of issues is the role of combinF3PY#9 complexity, an itsmeasuremen t,in determinFP the potential decisiveness of biometric decision making.

Over more than a half century of experience in research on learning, memory, and decision, I have come to believe that the most substantial and enduring advances have not been in the accumulation of empirical facts or the construction of models, but in the production of fruitful interactions between models and experimental research. Most experimental facts require continual reinterpretation and most models drop by the wayside like autumn leaves, but the results of interactions between models and experiments constitute most of our generalizable knowledge. Success in the interactive research effort depends not only on clearly formulated models and well-conducted experiments, but, just as importantly, on sound interpretations of the results of applying the models to the experiments. This interpretive phase of the effort is in some respects the most difficult, and I take as my main task in this article an account of some of the issues that have to be resolved and some of the traps that have to be avoided in order for the process to run to a successful conclusion. As a preliminary, I turn to a review of the basic concept of applying a model to data as it has evolved since its first rudimentary instantiation in the literature of memory and decision more than a century ago. Applying Models to Experiments Details of techniques for fitting curves, or, more broadly, formal models, whether mathematical or computer imple-This article presents in substance the author’s Governing Board Keynote

We describe a formal framework for analyzing how statistical properties of natural environments and the process of natural selection interact to determine the design of perceptual and cognitive systems. The framework consists of two parts: a Bayesian ideal observer with a utility function appropriate for natural selection, and a Bayesian formulation of Darwin’s theory of natural selection. Simulations of Bayesian natural selection were found to yield new insights, for example, into the co-evolution of camouflage, color vision, and decision criteria. The Bayesian framework captures and generalizes, in a formal way, many of the important ideas of other approaches to perception and cognition.

this paper and the writing of the paper were supported, in part, by grants to Dominic W. Massaro from the Public Health Service (PHS R01 NS 20314), the National Science Foundation (BNS 8812728), a James McKeen Cattell Fellowship, and the graduate division of the University of California, Santa Cruz. Cohen & Massaro On the Similarity of Categorization Models 2

This paper analyzes the issue of catastrophic fusion, a problem that occurs in multimodal recognition systems that integrate the output from several modules while working in non-stationary environments. For concreteness we frame the analysis with regard to the problem of automatic audio visual speech recognition (AVSR), but the issues at hand are very general and arise in multimodal recognition systems which need to work in a wide variety of contexts. Catastrophic fusion is said to have occurred when the performance of a multimodal system is inferior to the performance of some isolated modules, e.g., when the performance of the audio visual speech recognition system is inferior to that of the audio system alone. Catastrophic fusion arises because recognition modules make implicit assumptions and thus operate correctly only within a certain context. Practice shows that when modules are tested in contexts inconsistent with their assumptions, their influence on the fused product tends to increase, with catastrophic results. We propose a principled solution to this problem based upon Bayesian ideas of competitive models and inference robustification. We study the approach analytically on a classic Gaussian discrimination task and then apply it to a realistic problem on audio visual speech recognition (AVSR) with excellent results.

Speakers with a defective velopharyngeal mechanism produce speech with inappropriate nasal resonance (hypernasal speech). It is of clinical interest to detect hypernasality as it is indicative of an anatomical, neurological, or peripheral nervous system problem. There are various clinical techniques used to determine hypernasality. The current techniques are physically invasive or intrusive to some extent. A preferred approach for detecting hypernasality, would be noninvasive to maximize patient comfort and naturalness of speaking. In this study, a noninvasive technique based on the Teager Energy operator is proposed. Utilizing a property of the Teager Energy operator and a model for normal and nasalized speech, a significant difference between the Teager Energy profile for lowpass and bandpass filtered nasalized speech is shown. This difference is shown to be nonexistent for normal speech. A classification algorithm is formulated that detects the presence of hypernasality using a meas...

The function of vision is to get correct and useful answers about the state of the world. However, given that the state of the world is not uniquely specified by the visual input, the visual system must make good guesses or inferences. Thus, theories of visual system functions will be theories of inference, and we need a language in which theories of inference can be described. Analogous to calculus having a minimum expressiveness required to formulate theories in physics, we argue that the language of Bayesian inference is fundamental to quantitatively describe how reliable answers about the world can be obtained from image patterns. Bayes provides a minimal formalism that can deal with the sophistication and versatility of perception missing from some other approaches. Key missing components include the ability to model uncertainty, probabilistic modeling of pattern synthesis as a necessary prerequisite to understanding pattern inference, the means to handle the complexity of natural images, and the diversity of visual tasks. Most of the formal elements that we describe are not new and have their roots in signal detection theory and ideal observer analysis. We start from there to review and codify principles drawn from recent applications of Bayesian decision theory, Bayes nets and pattern theory to vision. To emphasize the