Amitava Dutta

Reacting to scheduling exceptions in FMS environments

Abstract Uncertainties in the production environment and modelling limitations inevitably result in operational deviations from schedules generated using predictive models. A production control mechanism monitors the environment for exceptions, and takes corrective actions, with the objective of adhering closely to planned objectives. This paper proposes a knowledge based (KB) methodology to perform such control in FMS environments. Use of KB techniques is motivated by the observation that control knowledge has a high heuristic content. A PROLOG implementation of this methodology, that generates automatic response to machine failures, dynamic introduction of new jobs, and dynamic increase in job priority, is presented. Experimental results appear to show that simple and generic design strategies for the KB can provide the basis for effective and robust control behavior. Handled by the Department of Computer and Information Sciences.

A multiperiod capacity planning model for backbone computer communication networks

The cost of transmission capacity constitutes a significant portion of the total investment cost of a backbone computer communications network. In this paper, we address the problem of deciding where, when and how much transmission capacity should be installed, over a multiperiod horizon, to meet increasing traffic requirements at minimum total discounted cost, while maintaining acceptable performance levels. The model allows traffic among existing nodes to increase, new nodes to be added to the network, and its topology to change, over time. It is formulated as an integer programming problem, and a Lagrangian relaxation based solution method is proposed. Capacity and routing decisions are made jointly over time in our multiperiod model. Furthermore, our method automatically provides numerical verification of solution quality through the Lagrangian lower bound. Computational experiments with several networks show that the method yields verifiably good solutions to this combinatorially explosive problem.

Integrating heuristic knowledge and optimization models for communication network design

A method is developed for integrating heuristic design knowledge with optimization models to create a tool for the topological design of computer communication networks. Design choices are based on suggestions from optimization models, as well as heuristic knowledge, which interact through a blackboard. A truth maintenance system (TMS) records justification for current design choices, as well as promising alternatives. A dependency-directed backtracking mechanism works with the TMS to choose other alternatives as warranted. This hybrid tool can consider a wider range of design requirements than is possible using one type of knowledge alone, is flexible in handling variations in these requirements, and has a modular structure which facilitates incremental refinement. Computational results on separate networks show it is effective in identifying good low-cost solutions. >

Reasoning with imprecise knowledge to enhance intelligent decision support

Decisionmakers often have to deal with knowledge that is both unstructured and imprecise in nature. Lack of structure forces the use of heuristic programming and artificial intelligence methods in automated decision support systems for such problems. At the same time, the knowledge representation and manipulation methods used should be able to include imprecise knowledge, where imprecision may be caused by uncertainty and/or fuzziness. The authors use their previously developed methodology (see Decision Support Syst. vol.2, no.3, 1986 and vol.2, no.4, 1986) for using imprecise reasoning in decision support systems to describe a prototype system based on this methodology. The systems behavior is examined using a set of sample problems. The knowledge representation mechanisms used are shown to have significant expressive power and facilitate the knowledge of engineering process through greater flexibility in the representation and use of imprecise knowledge. The system is able to provide additional information to the user that cannot readily be provided in systems limited to precise reasoning. >

Computer based support of reasoning in the presence of fuzziness

Abstract Many important problems encountered in managerial decision making are unstructured, making them difficult to solve by preset algorithms. Much of the complexity of such problems is due to the reasoning that is needed to construct solution procedures for each problem instance. Thus, any Decision Support System (DSS) designed for such problems should support this reasoning activity, in addition to data access and computational activities. Yet, existing DSS generally do not provide much reasoning support. One of the difficulties faced in building automated systems to support such reasoning is that much of the knowledge typically available for unstructured problems is imprecise, where imprecision may be caused by either fuzziness, uncertainty, or both. In this paper, we address the problem of supporting problem solving with fuzzy knowledge. We develop a formal method for representing fuzzy knowledge, using a framework of mathematical logic. Using this method, fuzziness in all the major constructs needed to describe knowledge can be represented. Relationships can be constructed using fuzzy operators and terms, and components in relationships can be weighted by their relative significance. Also, computational procedures and data access procedures can be directly integrated into the reasoning process. Knowledge thus represented can be manipulated using suitable reasoning mechanisms. The fuzzy inference methods we present, enable the generation of acceptable solutions to problems even when some of the knowledge used is highly imprecise and/or incomplete. Other desirable features, such as explanation of solution procedures and user participation in problem solving, are also supported by our methodology. In addition, we develop bounding procedures, which convey the imprecision in the reasoning process, and also help to reduce the complexity of the search process by pruning poor solutions. Finally, we describe a prototype system which implements the methods developed in this paper. Using example problems processed by the system, we illustrate the versatility of these methods, and also highlight their major features and potential utility in practical applications.

Computer based support of reasoning in the presence of fuzziness and uncertainty

Abstract Decision making tasks require that we access (possibly large volumes of) data, transform it in various ways by performing computations on it, as well as reason with such raw or transformed data. Also, imprecision is almost omnipresent in practical decision making environments. In order to provide computer based support for reasoning activities in such environments, it is therefore necessary to develop schemes to represent imprecise knowledge and mechanically manipulate it. A previous paper developed such a scheme for fuzzy knowledge. This paper extends that representation to capture uncertainty as well. Mechanisms for propagating uncertainty during the reasoning process are also developed. The propagation mechanism allows interaction between fuzziness and uncertainty during the course of reasoning. The scheme developed here has been tested on a prototype implementation. The robustness of the representation and its use are demonstrated with an example.

Design of private backbone networks -- I: time varying traffic

Abstract Technological advances, special communication needs and potential cost advantages are prompting an increasing number of business to configure private communication networks, bypassing the public networks to varying degrees. In view of the high prices of leasing capacity, this paper develops a model for determining transmission line capacity in private backbone networks, such that multi busy-hour point-to-point circuit requirements can be met at low cost. Recent technology has developed backbone modal elements, such as Networking Multiplexers/Digital Cross Connect Systems, which can be easily reconfigured to reroute traffic so as to take advantage of noncoincidence of demand among communicating node pairs across different busy-hours. A multi-period cost model is formulated. A 3-phase heuristic, based on Benders decomposition, is developed to determine the line capacities jointly with circuit routing to satisfy the time varying circuit requirements at low cost. Computational results on practical sized backbone private networks are shown and solution quality is established by comparison with a lower bound obtained from an LP relaxation of the model.

Design of private backbone networks -- II: time varying grouped traffic

Abstract In an earlier companion paper, we have developed a model to assign transmission capacities in private backbone networks, for multi busy-hour point-to-point circuit requirements. In this paper, we consider the situation when point-to-point demands between backbone nodes consist of one or more circuit groups , where each circuit group has one or more circuits. The granularity of routing is a circuit group. Grouping may arise due to multiplexing or heterogeniety of traffic types. The additional difficulties posed by grouping are (i) all circuits in a group are required to follow the same route during a busy-hour and (ii) circuits in the same group cannot be split across different high capacity lines even on the same edge of the backbone network. The 3-phase heuristic developed in the earlier paper is modified to accomodate grouping of circuits. Computational results on practical sizes are compared to a lower bound obtained from an LP relaxation of the problem.

The Explicit Support of Human Reasoning in Decision Support Systems

Publisher Summary The chapter highlights reasoning and its explicit support as a fundamental research issue in decision support system (DSS), together with additional developments in artificial intelligence (AI) that can be the basis for further theoretical advancements in decision support. DSS refers to computer-based systems that aid human decision-making activities in a variety of ways. Reasoning is a core activity in human decision-making processes. A well-known model of decision making consists of three phases—intelligence, design, and choice. The chapter discusses several reasons for the choice of managerial decision making as the type of human decision activity to be supported. It reviews some chronological developments in computer-based support of managerial decision-making, as it underscores the need for DSS to explicitly support reasoning activities. The chapter develops some representation techniques that support reasoning activities in DSS environments. The representation attempts to capture both fuzziness and uncertainty in the environment, and the manipulation methods propagate measures of imprecision, through a line of reasoning, to any conclusion that is drawn from a body of imprecise knowledge so represented.

AUTOREF: a deductive database for automatic referee selection

Abstract A central function in processing papers submitted for publication is that of finding suitable referees. Referees should be knowledgeable, have sufficient time, and be thorough and prompt in their reviews. Sometimes, specific knowledge, such as a known conflict of interest, will have a bearing on the process. While the record-keeping associated with submissions is often automated, the referee selection process is still done manually. We describe a prototype deductive database named AUTOREF, implemented in PROLOG, that can provide automated assistance in this selection process. AUTOREF strives to find the best referees, failing which, it progressively relaxes requirements to find the most suitable available ones. As a last resort, it goes into manual code. The editorial activities supported by AUTOREF are generic, and could be customized for most specific requirements. Automated assistance for referee selection might further improve quality and efficiency of the editorial process.