Generalizing the notion of schema in genetic algorithms

doi:10.1016/0004-3702(91)90019-G

Artificial Intelligence

Volume 50, Issue 3, August 1991, Pages 385-396

https://doi.org/10.1016/0004-3702(91)90019-G Get rights and content

Abstract

In this paper we examine some of the fundamental assumptions which are frequently used to explain the practical success which Genetic Algorithms (GAs) have enjoyed. Specifically, the concept of schema and the Schema Theorem are interpreted from a new perspective. This allows GAs to be regarded as a constrained random walk, and offers a view which is amenable to generalization. The minimal deceptive problem (a problem designed to mislead the genetic paradigm) is analyzed in the context provided by our interpretation, where a different aspect of its difficulty emerges.

References (2)

D.E. Goldberg
J.H. Holland

Cited by (78)

A hybrid particle swarm optimization and genetic algorithm for closed-loop supply chain network design in large-scale networks
2015, Applied Mathematical Modelling
Citation Excerpt :
In the 1990s, different aspects of the GA were extended; Hartl [39] researched the convergence of the algorithm, and Radcliffe [40] and Bean [41] studied the crossover operator (their studies are used here). Miller and Goldberg [49] studied the selection strategies, and Vose [50] and Koza [51] investigated the entire concept of the GA of Holland by proposing genetic programming. More information about genetic algorithm and other evolutionary algorithms can be found in Reeves [52], Mitchell [53], and Baeck’s et al. [54] studies.
Today, tracking the growing interest in closed-loop supply chain shown by both practitioners and academia is easily possible. There are many factors, which transform closed-loop supply chain issues into a unique and vital subject in supply chain management, such as environmental legislation, customer awareness, and the economical motivations of the organizations. However, designing and planning a closed-loop supply chain is an NP-hard problem, which makes it difficult to achieve acceptable results in a reasonable time. In this paper, we try to cope with this problem by proposing a new and effective solution methodology. On the other hand, this research considers improving closed-loop supply chain network optimization processes through dealing with mathematical programming tools; developing a deterministic multi-product, multi-echelon, multi-period model; and finally presenting an appropriate methodology to solve various sizes of instances. Both design and planning decision variables (location and allocation) are considered in the proposed network. Besides, in order to have a reliable performance evaluation process, large-scale instances are regarded in computational analysis. Two popular meta-heuristic algorithms are considered to develop a new elevated hybrid algorithm: the genetic algorithm (GA) and particle swarm optimization (PSO). Analyzing the above-mentioned algorithms’ strengths and weaknesses leads us to attempt to improve the GA using some aspects of PSO. Therefore, a new hybrid algorithm is proposed and a complete validation process is undertaken using CPLEX and MATLAB software. In small instances, the global optimum points of CPLEX for the proposed hybrid algorithm are compared to genetic algorithm, and particle swarm optimization. Then, in small, mid, and large-size instances, performances of the proposed meta-heuristics are analyzed and evaluated. Finally, a case study involving an Iranian hospital furniture manufacturer is used to evaluate the proposed solution approach. The results reveal the superiority of the proposed hybrid algorithm when compared to the GA and PSO.
Dynamic model of multi-agent social evolutionary algorithm and its convergence
2013, Journal of China Universities of Posts and Telecommunications
With a typical and simple 2-bit problem, a dynamic model of multi-agent social evolutionary algorithm (MASEA) is constructed by dynamic method. Then, the global dynamic shape of MASEA is comprehensively analyzed and the common evolution operators are also formally described. Furthermore, the effect that every evolutionary operator has on the dynamic shape is discovered by attraction analysis of the fixed points in the models. The global convergence of MASEA is also proved.
Redefinition of the KMV model's optimal default point based on genetic algorithms - Evidence from Taiwan
2011, Expert Systems with Applications
Citation Excerpt :
The so-called schema theorem shows that a genetic algorithm automatically allocates an exponentially increasing number of trials to the best observed schemata. This leads to a favorable trader between exploitation of promising directions of the search space and exploration of less-frequented regions of the space (see also Vose, 1991). However, there is no general result guaranteeing the convergence of a genetic algorithm to the global optimum.
In this paper, we propose a new method based on genetic algorithms to solve the optimal default point of the KMV model. In our empirical study, we compare the GA-KMV model with the QR-KMV and KMV models. The results indicate that the percentage of correctness of the GA-KMV model is higher than those for the other two models. This is to say, the GA-KMV model has a better goodness of fit. We also obtain the optimal default point for a Taiwan listed company. This can help us to predict the default point and improve the bank’s risk management performance.
Markowitz-based portfolio selection with minimum transaction lots, cardinality constraints and regarding sector capitalization using genetic algorithm
2009, Expert Systems with Applications
Heuristic algorithms strengthen researchers to solve more complex and combinatorial problems in a reasonable time. Markowitz’s Mean-Variance portfolio selection model is one of those aforesaid problems. Actually, Markowitz’s model is a nonlinear (quadratic) programming problem which has been solved by a variety of heuristic and non-heuristic techniques. In this paper a portfolio selection model which is based on Markowitz’s portfolio selection problem including three of the most important limitations is considered. The results can lead Markowitz’s model to a more practical one. Minimum transaction lots, cardinality constraints (both of which have been presented before in other researches) and market (sector) capitalization (which is proposed in this research for the first time as a constraint for Markowitz model), are considered in extended model. No study has ever proposed and solved this expanded model. To solve this mixed-integer nonlinear programming (NP-Hard), a corresponding genetic algorithm (GA) is utilized. Computational study is performed in two main parts; first, verifying and validating proposed GA and second, studying the applicability of presented model using large scale problems.
An analysis of the locality of binary representations in genetic and evolutionary algorithms
2021, PeerJ Computer Science
Soft computing in chemical and physical sciences: A shift in computing paradigm
2017, Soft Computing in Chemical and Physical Sciences: A Shift in Computing Paradigm

View all citing articles on Scopus

View full text

Research noteGeneralizing the notion of schema in genetic algorithms

Abstract

Research note
Generalizing the notion of schema in genetic algorithms