Entity

Time filter

Source Type


Zhan Z.-H.,Sun Yat Sen University | Zhang G.-Y.,Sun Yat Sen University | Ying-Lin,Key Laboratory Machine Intelligence and Advanced Computing | Ying-Lin,Sun Yat Sen University | Zhang J.,Key Laboratory Software Technology
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2014

This paper proposes to solve the task scheduling problem in cloud computing by using a load balance aware genetic algorithm (LAGA) with Minmin and Max-min methods. Task scheduling problems are of great importance in cloud computing, and become especially challenging when taking load balance into account. Our proposed LAGA algorithm has several advantages when solving this kind of problems. Firstly, by introducing the time load balance (TLB) model to help establish the fitness function with makespan, the algorithm benefits from the ability to find the solution that performs best on load balance among a set of solutions with the same makespan. More importantly, the interaction between makespan and TLB helps the algorithm to minimize makespan in the same time. Secondly, Min-min and Max-min methods are used to produce promising individuals at the beginning of evolution, leading to noticeable improvement of evolution efficiency. We evaluated LAGA on several task scheduling problems and compared with a Min-min, Max-min improved version of genetic algorithm (MMGA), which does not use the TLB strategy. The results show that LAGA can obtain very competitive results with good load balancing properties, and outperform MMGA in both makespan and TLB objectives. © Springer International Publishing Switzerland 2014. Source


Zhang M.-D.,Sun Yat Sen University | Zhang M.-D.,Key Laboratory Machine Intelligence and Advanced Computing | Zhang M.-D.,Key Laboratory Software Technology | Zhan Z.-H.,Sun Yat Sen University | And 6 more authors.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2014

Artificial bee colony (ABC) algorithm is a novel heuristic algorithm inspired from the intelligent behavior of honey bee swarm. ABC algorithm has a good performance on solving optimization problems of multivariable functions and has been applied in many fields. However, traditional ABC algorithm chooses solutions on the onlooker stage with roulette wheel selection (RWS) strategy which has several disadvantages. Firstly, RWS is suitable for maximization optimization problem. The fitness value has to be converted when solving minimization optimization problem. This makes RWS difficult to be generally used in real-world applications. Secondly, RWS has no any parameter that can control the selection pressure. Therefore, RWS is not easy to adapt to various optimization problems. This paper proposes a tournament selection based ABC (TSABC) algorithm to avoid these disadvantages of RWS based ABC. Moreover, this paper proposes an elitist strategy that can be applied to traditional ABC, TSABC, and any other ABC variants, so as to avoid the phenomenon that ABC algorithm may abandon the globally best solution in the scout stage. We compare the performance of traditional ABC and TSABC on a set of benchmark functions. The experiment results show that TSABC is more flexible and can be efficiently adapted to solve various optimization problems by controlling the selection pressure. © Springer International Publishing Switzerland 2014. Source


Jia Y.-H.,Sun Yat Sen University | Jia Y.-H.,Key Laboratory Machine Intelligence and Advanced Computing | Chen W.-N.,Sun Yat Sen University | Chen W.-N.,Key Laboratory Machine Intelligence and Advanced Computing | And 3 more authors.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2014

Search-based method using meta-heuristic algorithms is a hot topic in automatic test data generation. In this paper, we develop an automatic test data generating tool named particle swarm optimization data generation tool (PSODGT). The PSODGT is characterized by the following two features. First, the PSODGT adopts the condition-decision coverage (C/DC) as the criterion of software testing, aiming to build an efficient test data set that covers all conditions. Second, the PSODGT uses a particle swarm optimization (PSO) approach to generate test data set. In addition, a new position initialization technique is developed for PSO. Instead of initializing the test data randomly, the proposed technique uses the previously-found test data that can reach the target condition as the initial positions so that the search speed of PSODGT can be further accelerated. The PSODGT is tested on four practical programs. Experimental results show that the proposed PSO approach is promising. © Springer International Publishing Switzerland 2014. Source


Wang J.-B.,Sun Yat Sen University | Wang J.-B.,Key Laboratory Machine Intelligence and Advanced Computing | Chen W.-N.,Sun Yat Sen University | Chen W.-N.,Key Laboratory Machine Intelligence and Advanced Computing | And 4 more authors.
GECCO 2015 - Companion Publication of the 2015 Genetic and Evolutionary Computation Conference | Year: 2015

Portfolio optimization problems are challenging as they contain different kinds of constrains and their complexity becomes very high when the number of assets grows. In this paper, we develop a dimension-decreasing particle swarm optimization (DDPSO) for solving multi-constrained portfolio optimization problems. DDPSO improves the efficiency of PSO for solving portfolio optimization problems with a lot of asset and it can easily handle the cardinality constraint in portfolio optimization. To improve search diversity, the dimension-decreasing method is coupled with the comprehensive learning particle swarm optimization (CLPSO) algorithm. The proposed method is tested on benchmark problems from the OR library. Experimental results show that the proposed algorithm performs well. Copyright is held by the owner/author(s). Source


Yu X.,Sun Yat Sen University | Yu X.,Key Laboratory Machine Intelligence and Advanced Computing | Chen W.-N.,Sun Yat Sen University | Chen W.-N.,Key Laboratory Machine Intelligence and Advanced Computing | And 4 more authors.
GECCO 2015 - Proceedings of the 2015 Genetic and Evolutionary Computation Conference | Year: 2015

This paper takes the multiobjective traveling salesman problem (MOTSP) as the representative for multiobjective combinatorial problems and develop a set-based comprehensive learning particle swarm optimization (S-CLPSO) with decomposition for solving MOTSP. The main idea is to take advantages of both the multiobjective evolutionary algorithm based on decomposition (MOEA/D) framework and our previously proposed S-CLPSO method for discrete optimization. Consistent to MOEA/D, a multiobjective problem is decomposed into a set of subproblems, each of which is represented as a weight vector and solved by a particle. Thus the objective vector of a solution or the cost vector between two cities will be transformed into real fitness to be used in S-CLPSO for the exemplar construction, the heuristic information generation and the update of pBest. To validate the proposed method, experiments based on TSPLIB benchmark are conducted and the results indicate that the proposed algorithm can improve the solution quality to some degree. © 2015 ACM. Source

Discover hidden collaborations