An AHP model of DSS effectiveness

Empirical studies have reported equivocal, or even dysfunctional, results from the use of decision support systems (DSS). Recent examples are the Davis, Kottemann, and Remus production planning experiments. According to the researchers, these experiments demonstrate that DSS what-if analysis creates an ‘illusion of control’ that causes users to overestimate its effectiveness. Such experimental findings are contrary to case-supported DSS theory. This paper examines the discrepancy. It first overviews the decision-making process, presents a generic DSS, identifies the theoretical role of the DSS in improving decision making, develops a multiple criteria model of DSS effectiveness, and gives a DSS for delivering the model to users. Illustrating with recent empirical investigations and the Davis, Kottemann, and Remus studies, the DSS-delivered model is used to reconcile the incongruity between the experimental findings and the case-supported theory. The paper concludes with a discussion of the article's implications for information systems research and practice.     

Non-emergency patient transport services planning through genetic algorithms

Non-emergency Patient Transport Services (PTS) are provided by ambulance companies for patients who do not require urgent and emergency transport. These patients require transport to or from a health facility like a hospital, but due to clinical requirements are unable to use private or public transport. This task is performed nowadays mainly by human operators, spending a high amount of time and resources to obtain solutions that are suboptimal in most cases. To overcome this limitation, in this paper we present NURA (Non-Urgent transport Routing Algorithm), a novel algorithm aimed at ambulance route planning. In particular, NURA relies on a genetic algorithm to explore the solution space, and it includes a scheduling algorithm to generate detailed routes for ambulances. Experimental results show that NURA is able to outperform human experts in several real scenarios, reducing the time spent by patients in ambulances during non-emergency transportations, increasing ambulance usage, while saving time and money for ambulance companies.     

Capturing the high-risk environment of the transition economy in Bulgaria—a simulation-based DSS

Eastern European countries have undergone a transition from centralized economic planning to more open economic systems. Hard data based upon past experience are inappropriate for decision making in this radically changed environment. A team of Bulgarian and U.S. researchers utilized system dynamics simulation to model the microeconomic environment of a Bulgarian winery expanding into regional and international markets. Expert opinion was provided for both micro- and macroeconomic factors. Given the uncertainty of the data and the ambiguity in the experts' opinions, fuzzy logic was used to model the transitional economic firm's decision making.     

Extending NC and RNC algorithms

A technique is presented by which NC and RNC algorithms for some problems can be extended into NC and RNC algorithms, respectively, that solve more general parametric problems. The technique is demonstrated on explicit bounded degree circuits. Applications include parametric extensions of the shortest-path and spanning-tree problems and, in particular, the minimum-ratio-cycle problem, showing all these problems are in NC.     

Extending NC and RNC algorithms

A technique is presented by which NC and RNC algorithms for some problems can be extended into NC and RNC algorithms, respectively, that solve more general parametric problems. The technique is demonstrated on explicit bounded degree circuits. Applications include parametric extensions of the shortest-path and spanning-tree problems and, in particular, the minimum-ratio-cycle problem, showing all these problems are in NC.     

Placing actuators on space structures by genetic algorithms and effectiveness indices

Genetic algorithms are a powerful tool for the solution of combinatorial problems such as the actuator placement problem. However, they require a large number of analyses with correspondingly high computational costs. Therefore, it is useful to tune the operators and parameters of the algorithm on simple problems that are similar to more complex and computationally expensive problems. The present paper employs an easy-tocalculate measure of actuator effectiveness to evaluate several genetic algorithms. Additionally, the effects of population size and mutation rates are also investigated for a problem of placing actuators at 8 of 1507 possible locations. We find that even with the best of the algorithms and with optimum mutation rates, tens of thousands of analyses are required for obtaining near optimum locations. We propose a procedure that estimates the effectiveness of the various locations and discards ineffective ones, and find it helpful for reducing the cost of the genetic optimization.     

Improving multiple sequence alignment biological accuracy through genetic algorithms

Accuracy on multiple sequence alignments (MSA) is of great significance for such important biological applications as evolution and phylogenetic analysis, homology and domain structure prediction. In such analyses, alignment accuracy is crucial. In this paper, we investigate a combined scoring function capable of obtaining a good approximation to the biological quality of the alignment. The algorithm uses the information obtained by the different quality scores in order to improve the accuracy. The results show that the combined score is able to evaluate alignments better than the isolated scores.     

Flexible scheduling in a machining center through genetic algorithms

In this paper, as a practical application, we focus on a scheduling problem of a machining center which produces a variety of parts with a monthly processing plan. In order to take account of flexibility, such as partial troubles of a machining center, urgent orders and so forth, some parameters which reflect the decision maker's judgements for due-dates are introduced into objective functions of scheduling problems. Realizing that a direct application of conventional simple genetic algorithms to the formulated problems does not always give acceptable results, we introduce a genetic algorithm which is suitable for the formulated scheduling problems. Effectiveness of the proposed algorithm is demonstrated through some numerical simulations.     

Homogeneous genetic algorithms

A new type of genetic algorithm (GA) is developed to mitigate one or both of the following two major difficulties that traditional GAs may suffer: (1) when the number of ‘active genes’ needs to be held constant or kept within some prescribed range, and (2) when the set of genes is much larger than the set of active genes of feasible solutions under consideration. These homogeneous GAs use (unordered) sets to represent ‘active genes’ in chromosomes rather than strings, and a correspondingly natural crossover operator is introduced. ‘Homogeneous’ refers to the fact that, in contrast to traditional GAs where pairs of genes that are ‘close’ have better chances of being preserved under crossover, there is no notion of proximity between pairs of genes. Examples are provided that will demonstrate superior performance of these new GAs for some typical problems in which these difficulties arise.     

Convergence of genetic algorithms

The mean convergence of various versions of a genetic algorithm are considered. A number of convergence statements are formulated and relevant estimates are obtained. A hypothesis concerning the form of these estimates under variation of the structure of a genetic algorithm is put forward. Roman Riviyanovich Sharapov. Born 1981. Graduated from the Faculty of Mechanics and Mathematics, Ural State University, in 2003. Presently, a postgraduate at the Department of Mathematical Economics at the same university. Scientific interests: genetic algorithms, neural networks, and financial mathematics. Aleksandr Vyacheslavovich Lapshin. Born 1980. Graduated from the Faculty of Mechanics and Mathematics, Ural State University, in 2003. Presently, a postgraduate at the Department of Mathematical Economics at the same university. Scientific interests: financial mathematics, genetic algorithms, and neural networks.     

Leveraging saving-based algorithms by master-slave genetic algorithms

Saving-based algorithms are commonly used as inner mechanisms of efficient heuristic construction procedures. We present a general mechanism for enhancing the effectiveness of such heuristics based on a two-level genetic algorithm. The higher-level algorithm searches in the space of possible merge lists which are then used by the lower-level saving-based algorithm to build the solution. We describe the general framework and we illustrate its application to three hard combinatorial problems. Experimental results on three hard combinatorial optimization problems show that the approach is very effective and it enables considerable enhancement of the performance of saving-based algorithms.     

Acceleration of cellular automata algorithms using genetic algorithms

The following problem is solved: Given a Cellular Automaton with continuous state space which simulates a physical system or process, use a Genetic Algorithm in order to find a Cellular Automaton with discrete state space, having the smallest possible lattice size and the smallest possible number of discrete states, the results of which are as close as possible to the results of the Cellular Automaton with continuous state space. The Cellular Automaton with discrete state space evolves much faster than the Cellular Automaton with continuous state space. The state spaces of two Cellular Automata have been discretized using a Genetic Algorithm. The first Cellular Automaton simulates the two-dimensional photoresist etching process in integrated circuit fabrication and the second is used to predict forest fire spreading. A general method for the discretization of the state space of Cellular Automata using a Genetic Algorithm is also presented. The aim of this work is to provide a method for accelerating the execution of algorithms based on Cellular Automata (Cellular Automata algorithms) and to build a bridge between Cellular Automata as models for physical systems and processes and Cellular Automata as a VLSI architecture.     

Approximating the sheep milk production curve through the use of artificial neural networks and genetic algorithms

This paper examines the potential of a neural network coupled with genetic algorithms to recognize the parameters that define the production curve of sheep milk, in which production is time-dependent, using solely the data registered in the animals’ first controls. This enables the productive capacity of the animal to be identified more rapidly and leads to a faster selection process in determining the best producers. For this purpose we employ a network with a single hidden layer, using the property of “universal approximation”. To find the number of nodes to be included in this layer, genetic and pruning algorithms are applied. Results thus obtained applying genetic and pruning algorithms are found to be better than other models which exclusively apply the classical learning algorithm Extended-Delta-Bar-Delta.     

Entropy-Boltzmann selection in the genetic algorithms

A new selection method, entropy-Boltzmann selection, for genetic algorithms (GAs) is proposed. This selection method is based on entropy and importance sampling methods in Monte Carlo simulation. It naturally leads to adaptive fitness in which the fitness function does not stay fixed but varies with the environment. With the selection method, the algorithm can explore as many configurations as possible while exploiting better configurations, consequently helping to solve the premature convergence problem. To test the performance of the selection method, we use the NK-model and compared the performances of the proposed selection scheme with those of canonical GAs.     

遗传算法中的联结关系

进化计算领域的一个根本问题是哪些问题适合遗传算法求解,为此需要研究问题的结构对算法性能的影响.变量之间的联结关系是问题的本质属性,决定了遗传算法求解问题的难度.如果某个变量对函数值的影响非线性依赖于其他变量,则认为这些变量之间存的联结关系不,对遗传算法的联结关系这一理论问题进行了深入研究,给出了分析一般离散问题联结结构的理论基础,通过分析傅里叶系数与函数子空间的关系,提出了检测黑箱问题联结结构的确定性和随机性算法,通过试验分析说明了算法的正确性和有效性.     

Elitism-based compact genetic algorithms

This paper describes two elitism-based compact genetic algorithms (cGAs)-persistent elitist compact genetic algorithm (pe-cGA), and nonpersistent elitist compact genetic algorithm (ne-cGA). The aim is to design efficient cGAs by treating them as estimation of distribution algorithms (EDAs) for solving difficult optimization problems without compromising on memory and computation costs. The idea is to deal with issues connected with lack of memory by allowing a selection pressure that is high enough to offset the disruptive effect of uniform crossover. The pe-cGA finds a near optimal solution (i.e., a winner) that is maintained as long as other solutions generated from probability vectors are no better. The ne-cGA further improves the performance of the pe-cGA by avoiding strong elitism that may lead to premature convergence. It also maintains genetic diversity. This paper also proposes an analytic model for investigating convergence enhancement.     

General cardinality genetic algorithms

A complete generalization of the Vose genetic algorithm model from the binary to higher cardinality case is provided. Boolean AND and EXCLUSIVE-OR operators are replaced by multiplication and addition over rings of integers. Walsh matrices are generalized with finite Fourier transforms for higher cardinality usage. Comparison of results to the binary case are provided.     

On coevolutionary genetic algorithms

 The use of evolutionary computing techniques in coevolutionary/multi-agent systems is becoming increasingly popular. This paper presents simple models of the genetic algorithm in such systems, with the aim of examining the effects of different types of interdependence between individuals. Using the model it is shown that, for a fixed amount of interdependence between coevolving individuals, the existence of partner gene variance and the level at which fitness is applied can have significant effects, as does the evaluation partnering strategy used.     

Hierarchical distributed genetic algorithms

Genetic algorithm behavior is determined by the exploration/exploitation balance kept throughout the run. When this balance is disproportionate, the premature convergence problem will probably appear, causing a drop in the genetic algorithm's efficacy. One approach presented for dealing with this problem is the distributed genetic algorithm model. Its basic idea is to keep, in parallel, several subpopulations that are processed by genetic algorithms, with each one being independent from the others. Furthermore, a migration operator produces a chromosome exchange between the subpopulations. Making distinctions between the subpopulations of a distributed genetic algorithm by applying genetic algorithms with different configurations, we obtain the so-called heterogeneous distributed genetic algorithms. In this paper, we present a hierarchical model of distributed genetic algorithms in which a higher level distributed genetic algorithm joins different simple distributed genetic algorithms. Furthermore, with the union of the hierarchical structure presented and the idea of the heterogeneous distributed genetic algorithms, we propose a type of heterogeneous hierarchical distributed genetic algorithms, the hierarchical gradual distributed genetic algorithms. Experimental results show that the proposals consistently outperform equivalent sequential genetic algorithms and simple distributed genetic algorithms. ©1999 John Wiley & Sons, Inc.     

Where genetic algorithms excel

We analyze the performance of a genetic algorithm (GA) we call Culling, and a variety of other algorithms, on a problem we refer to as the Additive Search Problem (ASP). We show that the problem of learning the Ising perceptron is reducible to a noisy version of ASP. Noisy ASP is the first problem we are aware of where a genetic-type algorithm bests all known competitors. We generalize ASP to k-ASP to study whether GAs will achieve "implicit parallelism" in a problem with many more schemata. GAs fail to achieve this implicit parallelism, but we describe an algorithm we call Explicitly Parallel Search that succeeds. We also compute the optimal culling point for selective breeding, which turns out to be independent of the fitness function or the population distribution. We also analyze a mean field theoretic algorithm performing similarly to Culling on many problems. These results provide insight into when and how GAs can beat competing methods.