Abstract: In this paper, a new genetic algorithm for multi-objective optimization problems is introduced. That is called ”Neighborhood Cultivation GA (NCGA)”. In the recent studies such as SPEA2 or NSGA-II, it is demonstrated that some mechanisms are important; the mechanisms of placement in an archive of the excellent solutions, sharing without parameters, assign of fitness, selection and reflection the archived solutions to the search population. NCGA includes not only these mechanisms but also the neighborhood crossover. The comparison of NCGA with SPEA2 and NSGA-II by some test functions shows that NCGA is a robust algorithm to find Pareto-optimum solutions. Through the comparison between the case of using neighborhood crossover and the case of using normal crossover in NCGA, the effect of neighborhood crossover is made clear.

Abstract: Hyper-GA was introduced by the authors as a genetic algorithm based hyperheuristic which aims to evolve an ordering of low-level heuristics so as to find a good quality solution to a given problem. The adaptive length chromosome hyper-GA, let’s call it ALChyper-GA, is an extension of the authors previous work, in which the chromosome was of fixed length. The aim of a variable length chromosome is two fold; 1) it allows dynamic removal and insertion of heuristics 2) it allows the GA to find a good chromosome length which could otherwise only be found by experimentation. We apply the ALChyper-GA to a geographically distributed training staff and courses scheduling problem, and report that good quality solution can be found. We also present results for four versions of the ALChyper-GA, applied to five test data sets.

Abstract: In evolutionary algorithms based on probabilistic modeling, the offspring population is generated according to the estimated probability density model of the parent instead of using recombination and mutation operators. In this chapter, we have proposed a probabilistic model-building genetic algorithms (PMBGAs) for solving flow shop scheduling problems using edge histogram based sampling algorithms (EHBSAs). The effectiveness of introducing the tag node (TN) in a string representation is also discussed.

Abstract: We developed a multi-modal feed-forward neural network to predict the secondary structure of proteins. Several neural networks are used together and the final prediction results are decided by majority rule. We used 6137 residues to train and test the method. The average accuracy of the prediction is 66%, which is about 6.9% higher than single neural network.

Abstract: Parallelization of genetic algorithms (GAs) has received c onsiderable attention in recent years. The reason for this is the availability of su itable computational resources and the need for solving harder problems in reasonab le time. We describe a new parallel self-adaptive GA for solving the data cluster ing problem. The algorithm utilizes island parallelization implemented using g enebank model, in which GA processes communicate with each other only through the ge n bank process. This model allows one to easily implement different migrati on topologies. Experiments show that significant speedup can be reached by par allelization. The effect of migration parameters is also studied and the devel opment of diversity is examined by several measures, some of which are new.

Abstract: An interesting problem is under what circumstances will a collection of interacting agents realize efficient collective actions. This question will depend crucially on how self-interested agents interact and how they learn from each other. We model strategic interactions as dilemma games, coordination games or hawk-dove games. It is well known that the replicator dynamics based on natural selection converge to an inefficient equilibrium. In this paper, we focus on the effect of coevolutionary learning. Each agent is modeled to learn interaction rules defined as the function of own strategy and the strategy of the neighbor. We show that a collection of interacting agents converges into equilibrium in which the conditions of efficiency and equity are satisfied. We investigate interaction rules acquired by all agents and show that they share several rules with the common features to sustain equitable social efficiency. This paper also presents a comparative study of two evolving populations, one in a spatial environment, and the other in a small-world environment. The effect of the environment on the emergence of social efficiency is studied. The small-world environment is shown to encourage the emergence of social efficiency further than the spatial structure.

Abstract: In this paper the design and development of a hierarchical fuzzy logic Systems are investigated. A new method using genetic algorithms for design of hierarchical fuzzy logic systems are proposed. This research study is unique in the way proposed method is applied to design and development of hierarchical fuzzy logic systems. The proposed method is then applied to financial modelling and prediction. A hierarchical fuzzy logic system is developed to predict quarterly interest rates in Australia. The new method proposed determines the number of layer in a hierarchical fuzzy logic system. The advantages and disadvantages of using hierarchical fuzzy logic systems for financial modelling is also considered. Good prediction of quarterly interest rate in Australia is obtained using the above method. The number of fuzzy rule used are reduced dramatically and prediction of interest rate is improved.

Abstract: Optimization based on genetic algorithms is applied to the design of multilayered coatings, incorporating both coating-geometry and material-property optimization. The latter is based on parametric modeling of dielectric and magnetic properties of homogeneous materials, and effective-medium modeling of composites. Our approach treats physical laws to be obeyed by the models as constraints. Moreover, efficiency in thickness is considered in two ways: as a constraint on its upper limit, and in a multi-objective setting where we study both aggregation and Pareto optimality.

Abstract: This paper presents the use of multi-objective Genetic Algorithms (mGA) to solve the capacity and routing assignment problem arising in the design of self-healing networks using the Virtual Path (VP) concept. The aims to minimize the sum of working and backup capacity usage and transmission delay often compete and contradict with each other. Multi-objective Genetic algorithm is a powerful method for this kind of multi-objective problems. In this paper, a multi-objective GA approach is proposed to achieve the above two objectives while a set of customer traffic demands can still be satisfied and the traffic is 100% restorable under a single point of failure. We carried out a few experiments and the results illustrate the trade-off between objectives.

Abstract: In this paper, an integrated production and transportation scheduling model is proposed. This model is based on multi-item capacitated lot sizing and facility location type models. The objective of the integrated model is to minimize the total production and transportation cost. The integrated model is solved by Lagrangian decomposition method and the decomposed two sub-problems can be solved by genetic algorithm and Simplex method respectively. Computational results showed that the overall cost is reduced by 4% to 10% compared with the other two sequential optimization algorithms.

Abstract: The bandwidth-delay-constrained least-cost multicast routing is a challenging problem in high-speed multimedia networks. Computing such a constrained Steiner tree is an NP-complete problem. In this paper, we propose a novel QoS-based multicast routing algorithm based on the genetic algorithms (GA). In the proposed method, the predecessors encoding is used for genotype representation. Some novel heuristic algorithms are also proposed for mutation, crossover, and creation of random individuals. We evaluate the performance and efficiency of the proposed GA-based algorithm in comparison with other existing heuristic and GA-based algorithms by the result of simulation. This proposed algorithm has overcome all of the previous algorithms in the literatures.