Steven O. Kimbrough

Computers play the beer game: can artificial agents manage supply chains?

We model an electronic supply chain managed by artificial agents. We investigate whether artificial agents do better than humans when playing the MIT Beer Game. Can the artificial agents discover good and effective business strategies in supply chains both in stationary and non-stationary environments? Can the artificial agents discover policies that mitigate the Bullwhip effect? In particular, we study the following questions: Can agents learn reasonably good policies in the face of deterministic demand with fixed lead time? Can agents cope reasonably well in the face of stochastic demand with stochastic lead time? Can agents learn and adapt in various contexts to play the game? Can agents cooperate across the supply chain?

On automated message processing in electronic commerce and work support systems: speech act theory and expressive felicity

Electronic messaging, whether in an office environment or for electronic commerce, is normally carried out in natural language, even when supported by information systems. For a variety of reasons, it would be useful if electronic messaging systems could have semantic access to, that is, access to the meanings and contents of, the messages they process. Given that natural language understanding is not a practicable alternative, there remain three approaches to delivering systems with semantic access: electronic data interchange (EDI), tagged messages, and the development of a formal language for business communication (FLBC). We favor the latter approach. In this article we compare and contrast these three approaches, present a theoretical basis for an FLBC (using speech act theory), and describe a prototype implementation.

On a Feasible–Infeasible Two-Population (FI-2Pop) Genetic Algorithm for Constrained Optimization: Distance Tracing and no Free Lunch

We explore data-driven methods for gaining insight into the dynamics of a two-population genetic algorithm (GA), which has been effective in tests on constrained optimization problems. We track and compare one population of feasible solutions and another population of infeasible solutions. Feasible solutions are selected and bred to improve their objective function values. Infeasible solutions are selected and bred to reduce their constraint violations. Interbreeding between populations is completely indirect, that is, only through their offspring that happen to migrate to the other population. We introduce an empirical measure of distance, and apply it between individuals and between population centroids to monitor the progress of evolution. We find that the centroids of the two populations approach each other and stabilize. This is a valuable characterization of convergence. We find the infeasible population influences, and sometimes dominates, the genetic material of the optimum solution. Since the infeasible population is not evaluated by the objective function, it is free to explore boundary regions, where the optimum is likely to be found. Roughly speaking, the No Free Lunch theorems for optimization show that all blackbox algorithms (such as Genetic Algorithms) have the same average performance over the set of all problems. As such, our algorithm would, on average, be no better than random search or any other blackbox search method. However, we provide two general theorems that give conditions that render null the No Free Lunch results for the constrained optimization problem class we study. The approach taken here thereby escapes the No Free Lunch implications, per se.

Unique Names Violations, a Problem for Model Integration or You Say Tomato, I Say Tomahto

The tomato-tomahto problem (known as the synonymy problem in the database literature) arises in the context of model management when different names are used in different models for what should be identical variables, and these different models are to be integrated or combined into a larger model. When this problem occurs, it is said that the unique names assumption has been violated. We propose a method by which violations of the unique names assumption can be automatically detected. The method relies on declaring four kinds of information and modeling variables: dimensional information, laws relating dimensional expressions, information (called the quiddity) about the intended interpretation of the variables, and laws relating quiddity expressions. We present and discuss the method and the principles and theory behind it, and we describe our (prototype) implementation of the method, as an additional function of an existing model management system. INFORMS Journal on Computing, ISSN 1091-9856, was publis...

On automated discovery of models using genetic programming: bargaining in a three-agent coalitions game

The creation of mathematical, as well as qualitative (or rule-based), models is difficult, time-consuming, and expensive. Recent developments in evolutionary computation hold out the prospect that, for many problems of practical import, machine learning techniques can be used to discover useful models automatically. The prospects are particularly bright, we believe, for such automated discoveries in the context of game theory. This paper reports on a series of successful experiments in which we used a genetic programming regime to discover high-quality negotiation policies. The game-theoretic context in which we conducted these experiments-- a three-player coalitions game with sidepayments--is considerably more complex and subtle than any reported in the previous literature on machine learning applied to game theory.

Logic modeling: a tool for management science

Abstract Developments in logic and in information technology (especially the advent of logic programming) have converged to the point at which logic is, for a broad variety of problems, a useful tool to employ for modeling in areas of interest to management scientists. This paper presents the concept of logic modeling (model building with symbolic logic) and reviews several lines of research having in common a logic modeling approach to problems of interest in management scientist.

On Adaptive Emergence of Trust Behavior in the Game of Stag Hunt

We study the emergence of trust behavior at both the individual and the population levels. At the individual level, in contrast to prior research that views trust as fixed traits, we model the emergence of trust or cooperation as a result of trial and error learning by a computer algorithm borrowed from the field of artificial intelligence (Watkins 1989). We show that trust can indeed arise as a result of trial and error learning. Emergence of trust at the population level is modeled by a grid-world consisting of cells of individual agents, a technique known as spatialization in evolutionary game theory. We show that, under a wide range of assumptions, trusting individuals tend to take over the population and trust becomes a systematic property. At both individual and population levels, therefore, we argue that trust behaviors will often emerge as a result of learning.

On obligation, time, and defeasibility in systems for electronic commerce

A logical system is proposed for storing and reasoning about business messages. This system is envisioned to have mechanisms for reasoning defeasibly, reasoning about time, and reasoning about obligation. The authors describe how these capabilities allow the asking of questions that could otherwise not be asked about a message database. Further research opportunities are described.<<ETX>>

Formal language for business communication: sketch of a basic theory

Progress in research on developing a general-purpose formal language for business communications (FLBC) has been substantial. Key distinctions have been made, important concepts identified, and impressive prototypes implemented. The purpose of this paper is to present succinctly the basic elements of a complete (logical) theory for an FLBC. These elements include speech act theory, event semantics, extensive use of thematic roles, and representation in first-order logic. The theory is expected to lead to superior messaging in electronic commerce.

Exemplary documents: a foundation for information retrieval design

Documents are generally represented for retrieval by either extracting index terms from them or by creating and selecting from an external set of candidate terms. There are many procedures for doing this, but while work continues along these dimensions, there have been relatively few attempts to change this basic process. Of particular importance is the creation of indexing schemes for retrieval systems in nonlibrary contexts. Here, the cost of developing an indexing scheme independent of the documents to be retrieved is often considered too high to implement. As a result, simple full-text retrieval or, to a lesser extent, automatic extractive or associative indexing methods are the predominant methods used in nonlibrary contexts. This paper suggests an alternative document representation method based on what we call exemplary documents. Exemplary documents are those documents that describe or exhibit the intellectual structure of a particular field of interest. In so doing, they provide both an indexing vocabulary for that area and, more importantly, a narrative context in which the indexing terms have a clearer meaning. Further, it is much easier to develop an indexing scheme by using exemplary documents than it is to do so from scratch.