Vijay S. Mookerjee

Modeling Coordination in Software Construction: An Analytical Approach

Software development projects are typically team efforts, wherein groups of specialists work toward the common goal of building a software system. The individual efforts of team members need to be coordinated to ensure product quality and effectiveness of the team. In this paper we model the process of coordination in the construction phase of incrementally developed, modular software systems. The analytical model proposed here supports macro-level decisions regarding the development team size and the coordination policy, based upon micro-level interactions between the modules in a system. The objective in this model is to minimize the effort spent on coordination activities subject to the requirement that the system must be completed within a specified period. n nResults from the model are used to examine coordination related trade-offs. We show that: 1 more complex systems need a higher level of coordination than simpler ones, 2 if the time available for construction reduces, it is optimal to reduce the level of coordination, and 3 marginal productive output is a diminishing function of team size. The sensitivity of the analytical model with respect to its assumptions is studied by constructing a set of simulation experiments where these assumptions are relaxed. The results of these experiments provide support in establishing the robustness of the analytical model.

Inductive Expert System Design: Maximizing System Value

There is a growing interest in the use of induction to develop a special class of expert systems known as inductive expert systems. Existing approaches to develop inductive expert systems do not attempt to maximize system value and may therefore be of limited use to firms. We present an induction algorithm that seeks to develop inductive expert systems that maximize value. The task of developing an inductive expert system is looked upon as one of developing an optimal sequential information acquisition strategy. Information is acquired to reduce uncertainty only if the benefits gained from acquiring the information exceed its cost. Existing approaches ignore the costs and benefits of acquiring information. We compare the systems developed by our algorithm with those developed by the popular ID3 algorithm. In addition, we present results from an extensive set of experiments that indicate that our algorithm will result in more valuable systems than the ID3 algorithm and the ID3 algorithm with pessimistic pruning.

Sequential decision models for expert system optimization

Sequential decision models are an important element of expert system optimization when the cost or time to collect inputs is significant and inputs are not known until the system operates. Many expert systems in business, engineering, and medicine have benefited from sequential decision technology. In this survey, we unify the disparate literature on sequential decision models to improve comprehensibility and accessibility. We separate formulation of sequential decision models from solution techniques. For model formulation, we classify sequential decision models by objective (cost minimization versus value maximization) knowledge source (rules, data, belief network, etc.), and optimized form (decision tree, path, input order). A wide variety of sequential decision models are discussed in this taxonomy. For solution techniques, we demonstrate how search methods and heuristics are influenced by economic objective, knowledge source, and optimized form. We discuss open research problems to stimulate additional research and development.

Purchasing demand information in a stochastic-demand inventory system

This paper studies a periodic-review, stochastic-demand inventory system in which the manager has the opportunity each period to purchase information about demand in the upcoming period before deciding how much product to order. We analyze the information-purchase and product-replenishment decisions for both perfect and imperfect demand information. Under perfect information, we provide a characterization of the optimal policy for both finite and infinite horizon problems, and also establish useful managerial insights into the behavior of the system. We show that future demand information becomes less valuable at higher inventory levels, and more valuable when longer horizons remain. When the initial inventory is zero, solving the perfect-information problem reduces to computing a single quantity, for which we provide a closed form expression. As a result, this problem is shown to be equivalent to one in which the manager purchases perfect information over the entire horizon with a single lump-sum payment at the beginning of the horizon. Our analytical and numerical results demonstrate that most of the insights from the perfect-information scenario carry over to the imperfect-information case.

Improving the Performance Stability of Inductive Expert Systems Under Input Noise

Inductive expert systems typically operate with imperfect or noisy input attributes. We study design differences in inductive expert systems arising from implicit versus explicit handling of input noise. Most previous approaches use an implicit approach wherein inductive expert systems are constructed using input data of quality comparable to problems the system will be called upon to solve. We develop an explicit algorithm (ID3ecp) that uses a clean (without input errors) training set and an explicit measure of the input noise level and compare it to a traditional implicit algorithm, ID3p (the ID3 algorithm with the pessimistic pruning procedure). The novel feature of the explicit algorithm is that it injects noise in a controlled rather than random manner in order to reduce the performance variance due to noise. We show analytically that the implicit algorithm has the same expected partitioning behavior as the explicit algorithm. In contrast, however, the partitioning behavior of the explicit algorithm ...

Redesigning Case Retrieval to Reduce Information Acquisition Costs

Retrieval of a set of cases similar to a new case is a problem common to a number of machine learning approaches such as nearest neighbor algorithms, conceptual clustering, and case based reasoning. A limitation of most case retrieval algorithms is their lack of attention to information acquisition costs. When information acquisition costs are considered, cost reduction is hampered by the practice of separating concept formation and retrieval strategy formation. n nTo demonstrate the above claim, we examine two approaches. The first approach separates concept formation and retrieval strategy formation. To form a retrieval strategy in this approach, we develop the CRlc case retrieval loss criterion algorithm that selects attributes in ascending order of expected loss. The second approach jointly optimizes concept formation and retrieval strategy formation using a cost based variant of the ID3 algorithm ID3c. ID3c builds a decision tree wherein attributes are selected using entropy reduction per unit information acquisition cost. n nExperiments with four data sets are described in which algorithm, attribute cost coefficient of variation, and matching threshold are factors. The experimental results demonstrate that i jointly optimizing concept formation and retrieval strategy formation has substantial benefits, and ii using cost considerations can significantly reduce information acquisition costs, even if concept formation and retrieval strategy formation are separated.

Mean-Risk Trade-Offs in Inductive Expert Systems

Notably absent in previous research on inductive expert systems is the study of meanrisk trade-offs. Such trade-offs may be significant when there are asymmetries such as unequal classification costs, and uncertainties in classification and information acquisition costs. The objective of this research is to developmodels to evaluate mean-risk trade-offs in value-based inductive approaches. We develop a combined mean-risk measure and incorporate it into the Risk-Based induction algorithm. The mean-risk measure has desirable theoretical properties (consistency and separability) and is supported by empirical results on decision making under risk. Simulation results using the Risk-Based algorithm demonstrate: (i) an order of magnitude performance difference between mean-based and risk-based algorithms and (ii) an increase in the performance difference between these algorithms as either risk aversion, uncertainty, or asymmetry increases given modest thresholds of the other two factors.

Optimizing Expert Systems: Heuristics for Efficiently Generating Low-Cost Information Acquisition Strategies

We study the sequential information acquisition problem for rule-based expert systems as follows: find the information acquisition strategy that minimizes the expected cost to operate the system while maintaining the same output decisions. This problem arises for rule-based expert systems when the cost or time to collect inputs is significant and the inputs are not known until the system operates. We develop several optimistic heuristics to generate information acquisition strategies and study their properties. The heuristics provide choices concerning precision (i.e., how optimistic) versus computational effort. The heuristics are embedded into an informed search algorithm (based on AO*) that produces an optimal strategy and a greedy search algorithm. The search strategies are designed for situations in which rules can overlap and the cost of collecting an input may depend on the set of inputs previously collected. We study the properties of these approaches and simulate their performance on a variety of synthetic expert systems. Our results indicate that the heuristics are very precise, leading to near optimal results for greedy search and moderate search effort for optimal search.

Expert system design: minimizing information acquisition costs

Abstract Today, many organizations are investing heavily in expert systems. Unfortunately, many of these systems will fail to deliver the maximum possible value to their investors because little attention has been paid to the cost of providing these systems with the information they require to make a decision. In an expert system, the cost of providing the information that the system requires can be substantial. Minimizing information costs without affecting the decisions made by the system can reduce the cost of operating the system and thereby increase value. We develop an algorithm that determines an optimal information acquisition strategy for an existing system and show how a specific information acquisition strategy can be implemented. Because of the computational complexity of the algorithm, we also develop a simpler, heuristic solution to the problem. Our tests indicate that the heuristic performs very well. Prolog implementations for the same problem, on the other hand, perform poorly.

An economic approach to the development of inductive expert systems

Recently, there has been growing interest in the use of machine induction to develop expert systems. This approach offers an alternative to the costly and laborious process of manually extracting human knowledge to develop expert systems. In spite of the increasing commercial interest in inductive expert systems, the approaches used seldom attempt to maximize the value of the system. The authors present an algorithm that develops an inductive expert system with the objective of maximizing system value. They compare the performance of their algorithm to that of the popular ID3 algorithm. In their study, the decision trees produced by their algorithm were able to perform classification tasks at lower cost and were at least as accurate as the decision trees produced by ID3. In addition, their algorithm produced much smaller decision trees than those produced by ID3.<<ETX>>