Cliff T. Ragsdale

Combining a neural network with a genetic algorithm for process parameter optimization

Abstract A neural-network model has been developed to predict the value of a critical strength parameter (internal bond) in a particleboard manufacturing process, based on process operating parameters and conditions. A genetic algorithm was then applied to the trained neural network model to determine the process parameter values that would result in desired levels of the strength parameter for given operating conditions. The integrated NN–GA system was successful in determining the process parameter values needed under different conditions, and at various stages in the process, to provide the desired level of internal bond. The NN–GA tool allows a manufacturer to quickly determine the values of critical process parameters needed to achieve acceptable levels of board strength, based on current operating conditions and the stage of manufacturing.

A new approach to solving the multiple traveling salesperson problem using genetic algorithms

The multiple traveling salesperson problem (MTSP) involves scheduling m > 1 salespersons to visit a set of n > m locations so that each location is visited exactly once while minimizing the total (or maximum) distance traveled by the salespersons. The MTSP is similar to the notoriously difficult traveling salesperson problem (TSP) with the added complication that each location may be visited by any one of the salespersons. Previous studies investigated solving the MTSP with genetic algorithms (GAs) using standard TSP chromosomes and operators. This paper proposes a new GA chromosome and related operators for the MTSP and compares the theoretical properties and computational performance of the proposed technique to previous work. Computational testing shows the new approach results in a smaller search space and, in many cases, produces better solutions than previous techniques.

The Competitive Market Efficiency of Hotel Brands: An Application of Data Envelopment Analysis:

The objective of this article is to illustrate how managers in the hotel industry can analyze and improve their brands’market efficiency using data envelopment analysis (DEA). The authors evaluated the competitive market efficiency of 46 hotel brands in terms of customer satisfaction and customer value. Their DEA results show that 23 of the 46 brands studied generated less customer satisfaction and customer value for the same level of inputs relative to their more efficient competitors. In particular, the competitive market-inefficient hotel brands suffered more guest complaints, employed a lower quality staff, did not maintain their properties as well, and charged prices higher than justified by their market offerings. Furthermore, the competitive market-inefficient hotels had a less-than-optimal number of rooms and properties in the chain. The results illustrate how managers can use DEA to improve the relative market efficiency of their brands.

A Simulated Annealing Genetic Algorithm for the Electrical Power Districting Problem

Due to a variety of political, economic, and technological factors, many national electricity industries around the globe are transforming from non-competitive monopolies with centralized systems to decentralized operations with competitive business units. A key challenge faced by energy restructuring specialists at the World Bank is trying to simultaneously optimize the various criteria one can use to judge the fairness and commercial viability of a particular power districting plan. This research introduces and tests a new algorithm for solving the electrical power districting problem in the context of the Republic of Ghana and using a random test problem generator. We show that our mimetic algorithm, the Simulated Annealing Genetic Algorithm, outperforms a well-known Parallel Simulated Annealing heuristic on this new and interesting problem manifested by the deregulation of electricity markets.

On the classification gap in mathematical programming-based approaches to the discriminant problem

This article proposes a mathematical-programming-based approach to solve the classification problem in discriminant analysis which explicitly considers the classification gap. The procedure consists of two distinct phases and initially treats the classification gap as a fuzzy set in which the classification rule is not yet established. The nature of the classification gap is examined and a variety of methods are discussed which can be applied to identify the most appropriate classification rule over the fuzzy set. The proposed methodology has several potential advantages. First, it offers a more refined approach to the classification problem, facilitating careful analysis of the fuzzy region where the classification decision may not be obvious. Secondly, the two-phase approach enables the analysis of larger data sets when using computer-intensive procedures such as mixed-integer programming. Finally, because of the restricted choice of separating hyperplanes in phase 2, the approach appears to be more robust than other classification techniques with respect to outlier-contaminated data conditions. The robustness issue and computational advantage of our proposed methodology are illustrated using a limited simulation experiment.

Scheduling pre-printed newspaper advertising inserts using genetic algorithms

In recent years, the use of pre-printed advertising inserts in newspapers has increased dramatically. Pre-printed inserts allow advertisers to deliver colorful, high-quality marketing material to targeted groups of consumers within the newspapers delivery zone structure. To accommodate the increased workload associated with pre-printed inserts without negatively impacting the news deadline or delivery schedules, many newspaper companies face increasingly complex post-press scheduling decisions. This paper presents a spreadsheet model developed to represent the pre-printed insert scheduling problem in a case study of an actual medium-size newspaper company. The performance of two commercial genetic algorithm (GA) optimizers is compared on this problem. Computational testing shows the GAs develop schedules that substantially reduce the post-press production departments insert processing time.

Modeling optimization problems in the unstructured world of spreadsheets

Electronic spreadsheets are the most common software tool managers use to analyze data and model quantitative problems. Increasingly, these software packages are being used in introductory OR/MS courses to introduce students to a variety of quantitative modeling tools. Because spreadsheets are inherently free-form, they impose no particular guidelines or structure on the way problems may be modeled. Thus, academics and practitioners accustomed to solving problems using very structured, dedicated OR/MS software packages are facing the challenge of dealing with these problems in the unstructured spreadsheet environment where there is often a variety of ways to implement and solve the same problem. This challenge is particularly acute in the case of optimization problems. Some are responding to this challenge by devising rules for implementing models that impose an artificial structure on spreadsheets, sometimes resembling the operation of dedicated OR/MS optimization packages. This paper offers a critique of this approach and provides some guidelines we believe to be more helpful in creating effective spreadsheet models for optimization problems.

A spreadsheet-based decision support system for wood panel manufacturing

Abstract Wood paneling manufacturers face a number of complex decisions when trying to allocate production resources and combine various raw materials to meet production goals. While various linear programming formulations for this problem have been proposed, these models are often difficult to use and maintain for real-time decision making in a dynamic shop floor environment. This paper describes an MS Excel-based decision support system for wood panel manufacturing. The system is easy to use and maintain yet gives shop floor personnel access to powerful optimization capabilities useful for fine-tuning production processes in the face of changing supply and price situations.

A GROUPING GENETIC ALGORITHM FOR THE MULTIPLE TRAVELING SALESPERSON PROBLEM

The multiple traveling salesperson problem (MTSP) involves scheduling m > 1 salespersons to visit a set of n > m locations. Thus, the n locations must be divided into m groups and arranged so that each salesperson has an ordered set of cities to visit. The grouping genetic algorithm (GGA) is a type of genetic algorithm (GA) designed particularly for grouping problems. It has been successfully applied to a variety of grouping problems. This paper focuses on the application of a GGA to solve the MTSP. Our GGA introduces a new chromosome representation to indicate which salesperson is assigned to each tour and the ordering of the cities within each tour. We compare our method to standard GAs that employ either the one-chromosome or two-chromosome representation for MTSP. This research demonstrates that our GGA with its new chromosome representation is capable of solving a variety of MTSP problems from the literature and can outperform the traditional encodings of previously published GA methods.

A decision support methodology for stochastic multi-criteria linear programming using spreadsheets

In recent years, tools for solving optimization problems have become widely available through the integration of optimization software (or solvers) with all major spreadsheet packages. These solvers are highly effective on traditional linear programming (LP) problems with known, deterministic parameters. However, thoughtful analysts may rightly question the quality and robustness of optimal solutions to problems where point estimates are substituted for model parameters that are stochastic in nature. Additionally, while many LP problems implicitly involve multiple objectives, current spreadsheet solvers provide no convenient facility for dealing with more than one objective. This paper introduces a decision support methodology for identifying robust solutions to LP problems involving stochastic parameters and multiple criteria using spreadsheets.