George F. Luger

The architecture of a network level intrusion detection system

This paper presents the preliminary architecture of a network level intrusion detection system. The proposed system will monitor base level information in network packets (source, destination, packet size, and time), learning the normal patterns and announcing anomalies as they occur. The goal of this research is to determine the applicability of current intrusion detection technology to the detection of network level intrusions. In particular, the authors are investigating the possibility of using this technology to detect and react to worm programs.

Transfer effects in isomorphic problem situations

Abstract The results of a recent experiment to assess the relationship between human problem solving behaviour and the structural properties of certain problems are presented. The results suggest that human performance in a problem-solving task is significantly influenced by immediately prior experience on a different but structurally identical problem. Specifically, transfer effects are demonstrated across two problems of isomorphic structure.

Mathematical Model Building in the Solution of Mechanics Problems: Human Protocols and the MECHO Trace

This paper describes model building and manipulation in the solution of problems in mechanics. An automatic problem solver, MECHO, solving problems in several areas of mechanics, employs (1) a knowledge base representing the semantic content of the particular problem area, (2) a means-ends search strategy similar to GPS to produce sets of simultaneous equations and (3) a “focusing” technique, based on the data within the knowledge base, to guide the GSP-like search through possible equation instantiations. Sets of predicate logic statements are employed to describe this model building activity. These clauses are used to give information content to the knowledge base and provide both basis and guidance for the goal driven search. It is hypothesized that human subjects solving mechanics problems employ similar model building techniques. Protocols of several subjects are presented and comparisons are drawn with the traces of the automated problem solver. Adjustments are made to the program to provide a better fit of traces to protocols. Some implications are presented for using a rule based model for describing human problem solving performance in solving mechanics problems.

SHOCK PHYSICS DATA RECONSTRUCTION USING SUPPORT VECTOR REGRESSION

This paper considers a set of shock physics experiments that investigate how materials respond to the extremes of deformation, pressure, and temperature when exposed to shock waves. Due to the complexity and the cost of these tests, the available experimental data set is often very sparse. A support vector machine (SVM) technique for regression is used for data estimation of velocity measurements from the underlying experiments. Because of good generalization performance, the SVM method successfully interpolates the experimental data. The analysis of the resulting velocity surface provides more information on the physical phenomena of the experiment. Additionally, the estimated data can be used to identify outlier data sets, as well as to increase the understanding of the other data from the experiment.

A Model of the Development of the Early Infant Object Concept

A computational model is proposed for the early stages of development of the object concept in infancy. Stages in development are represented as a sequence of grammars or rewrite rules that parse a set of perceptual phenomena. The infants changes between developmental stages can be described by differences between the grammar rules that model each stage. The program replicates five studies by Bower et al on the development of the object concept and reaffirms the primacy of rest and motion parameters as explanatory invariants in early object-concept development.

Automated annotation of functional imaging experiments via multi-label classification.

Identifying the experimental methods in human neuroimaging papers is important for grouping meaningfully similar experiments for meta-analyses. Currently, this can only be done by human readers. We present the performance of common machine learning (text mining) methods applied to the problem of automatically classifying or labeling this literature. Labeling terms are from the Cognitive Paradigm Ontology (CogPO), the text corpora are abstracts of published functional neuroimaging papers, and the methods use the performance of a human expert as training data. We aim to replicate the experts annotation of multiple labels per abstract identifying the experimental stimuli, cognitive paradigms, response types, and other relevant dimensions of the experiments. We use several standard machine learning methods: naive Bayes (NB), k-nearest neighbor, and support vector machines (specifically SMO or sequential minimal optimization). Exact match performance ranged from only 15% in the worst cases to 78% in the best cases. NB methods combined with binary relevance transformations performed strongly and were robust to overfitting. This collection of results demonstrates what can be achieved with off-the-shelf software components and little to no pre-processing of raw text.

Using an expert system to explore enhanced oil recovery methods

Abstract This paper describes the use of an expert system, written with inexpensive shells (CLIPS and EXSHELL) for running on personal computers (PCs), to assist in selecting complex petroleum recovery processes. CLIPS is a forward-chaining rule-based system written in C, with rules entered in a LISP-like format. EXSHELL is a backward-chaining rule-based system written in PROLOG. These shells were used to write a system, an expert assistant, for use by petroleum engineers to screen candidate processes for enhanced oil recovery (EOR). The final choice is always made on the basis of economic evaluations. Testing has shown that the expert assistant greatly reduces the amount of work involved in making this choice. Rather than doing exhaustive economic calculations for all possible processes, the work is reduced to an economic comparison between the two or three most promising candidates. Rather than having to glean information and data from graphs or charts in technical papers, the user and the system work interactively to obtain the needed information. The system automatically selects the optimal paths to the solutions and is easily updated as new data on recovery processes become available. This paper also demonstrates the utility and power of these inexpensive shells, compares the approach used by each, and demonstrates the relative advantages of data-driven vs goal-driven search for this screening problem.

Modelling the stages of the identity theory of object-concept development in infancy.

A computational model is presented for the three stages of development of the object concept in infancy identified by Bower and Wishart in their research. The stages are described by sets of PROLOG clauses that interpret object structures representing the perceptual phenomena interpreted by the infants themselves. The infants changes between developmental stages can be described by differences between the rules modelling each stage. Three experiments are presented and the behaviour of the PROLOG model is described for each stage of development. Motion, rest, and boundedness of objects constitute the theoretical underpinning of the running PROLOG model and are hypothesized as the invariant aspects of perception that explain the behaviour of the infant at each stage of development. A possible explanation for transitions between stages is offered and justified in part by the output of the model, which in turn is used to predict the behavioural outcome of an experiment.

Designing a portable architecture for intelligent particle accelerator control

We present a portable system for intelligent control of particle accelerators. This system is based on a hierarchical distributed architecture. At the lowest level, a physical access layer provides an object-oriented abstraction of the target system. A series of intermediate layers implement general algorithms for control, optimization, data interpretation, and diagnosis. Decision making and planning are organized by knowledge-based components that utilize knowledge acquired from human experts to appropriately direct and configure lower level services. The general nature of the representations and algorithms at lower levels gives this architecture a high degree of potential portability. The knowledge-based decision-making and planning at higher levels gives this system an adaptive capability as well as making it readily configurable to new environments.

Representing Diversity in Communities of Bayesian Decision-makers

High-quality information has emerged from the contributions of many using the wiki paradigm. A logical next step is to use the wisdom of the crowd philosophy to solve complex problems and produce informed policy. We introduce a new approach to aggregating the beliefs and preferences of many individuals to form models that can be used in social policy and decision-making. Traditional social choice functions used to aggregate beliefs and preferences attempt to find a single solution for the whole population, but may produce an irrational social choice when a stalemate between opposing objectives occurs. Our approach, called collective belief aggregation, partitions a population into collectives that share a preference order over the expected utilities of decision options or the posterior likelihoods of a probabilistic variable. It can be shown that if a group of individuals share a preference order over the options, their aggregate will uphold principles of rational aggregation defined by social choice theorists. Super-agents can then be formed for each collective that accurately represent the preferences of their collective. These super-agents can be used to represent the collectives in decision analysis and decision-making tasks. We demonstrate the potential of using collective belief aggregation to incorporate the objectives of stakeholders in policy-making using preferences elicited from people about healthcare policy.