Yongsheng Liang

Inferior solutions in Gaussian EDA: Useless or useful?

Estimation of distribution algorithms (EDAs) are a special class of model-based evolutionary algorithms (EAs). To improve the performance of traditional EDAs, many remedies were suggested, which mainly focused on estimating a suitable probability distribution model with superior solutions. Different from existing research ideas, this paper tries to enhance EDA by exploiting the potential value of inferior solutions, where Gaussian EDA is taken as an example. It will be shown that, after a simple repair operation, inferior solutions could be surprisingly useful in adjusting the covariance matrix of Gaussian model, then a better search direction and a more proper search scale can be obtained. Since the aim of Inferior Solution Repairing (ISR) operator is not to directly improve the quality of inferior solutions, but to make them closer to superior ones, it can be implemented in a simple way. Combining ISR and traditional Gaussian EDA, a new EDA variant named ISR-EDA is developed. Comparison with existing EDAs and some other state-of-the-art EAs on benchmark functions demonstrates that ISR-EDA is efficient and competitive.

Enhance continuous estimation of distribution algorithm by variance enlargement and reflecting sampling

Estimation of distribution algorithm (EDA) is a kind of typical model-based evolutionary algorithm (EA). Although possessing competitive advantages in theoretical analysis, current EDAs may encounter premature convergence due to the rapid shrinkage of the search range and the relatively low sampling efficiency. Focusing on continuous EDAs with Gaussian models, this paper proposes a novel probability density estimator which can adaptively enlarge the variances and thus endow EDA with flexible search behavior. For the estimated probability density, a reflecting sampling strategy which can further improve the search efficiency is put forward. With these two algorithmic strategies, a new EDA variant named EDAver is developed. Experimental results on a set of benchmark problems demonstrate that EDAver outperforms conventional EDAs and can produce superior solutions in comparison with some state-of-the-art EAs.

Improved invasive weed optimization based on clustering strategy

Invasive weed optimization (IWO), which is inspired from the invasive behavior of weeds growth in nature, is a population-based intelligence algorithm. However, competitive exclusion may shrink search space and place most seeds in the same local area. Meanwhile, the accurate value of standard deviation is not easy to determine. These two shortcomings may lead to premature convergence and unable to achieve the global optimum. In order to overcome these two shortcomings, a clustering IWO (CIWO) is proposed by incorporating the core idea of clustering into IWO. We introduce a clustering strategy which is deployed before reproduction to disperse solution regions so that new seeds can locate in different areas. In addition, the value of standard deviation is based on statistical information and calculated from fittest individuals of each cluster so they can be accurate enough to the actual value and more representative. We compare it with the basic IWO, and a modified particle swarm optimization on a set of 14 benchmark functions. Experimental results indicate that CIWO is an effective and efficient algorithm can not only obtain the result superior to the standard invasive weed optimization but also explores and exploits the promising regions in the search space effectively.

Niching an archive-based gaussian estimation of distribution algorithm via adaptive clustering

Traditional Gaussian estimation of distribution algorithm (EDA) may suffer from premature convergence and has a high risk of falling into local optimum when dealing with multimodal problem. In this paper, we first attempt to improve the performance of EDA by utilizing historical solutions and develop a novel archive-based EDA variant. The use of historical solutions not only enhances the search efficiency of EDA to a large extent, but also significantly reduces the population size so that a faster convergence could be achieved. Then, the archive-based EDA is further integrated with an novel adaptive clustering strategy for solving multimodal optimization problems. Taking the advantage of the clustering strategy in locating different promising areas and the powerful exploitation ability of the archive-based EDA, the resultant algorithm is endowed with strong capability in finding multiple optima. To verify the efficiency of the proposed algorithm, we tested it on a set of niching benchmark problems, the experimental results indicate that the proposed algorithm is competitive.

Enhancing cooperative coevolution for large scale optimization by adaptively constructing surrogate models

It has been shown that cooperative coevolution (CC) can effectively deal with large scale optimization problems (LSOPs) through a divide-and-conquer strategy. However, its performance is severely restricted by the current context-vector-based sub-solution evaluation method because this method requires too many computation resources. To alleviate this issue, this study proposes an adaptive surrogate model assisted CC framework which adaptively constructs surrogate models for different sub-problems by fully considering their characteristics. By this means, the computation cost could be greatly reduced without significantly sacrificing evaluation quality. Empirical studies on IEEE CEC 2010 large scale benchmark suit show that the concrete algorithm based on this framework performs well.

A historical interdependency based differential grouping algorithm for large scale global optimization

Cooperative co-evolution (CC) is a powerful evolutionary computation framework for solving large scale global optimization (LSGO) problems via the strategy of divide-and-conquer, but its efficiency highly relies on the decomposition result. Existing decomposition algorithms either cannot obtain correct decomposition results or require a large number of fitness evaluations (FEs). To alleviate these limitations, this paper proposes a new decomposition algorithm named historical interdependency based differential grouping (HIDG). HIDG detects interdependency from the perspective of vectors. By utilizing historical interdependency information, it develops a novel criterion which can directly deduce the interdependencies among some vectors without consuming extra FEs. Coupled with an existing vector-based decomposition framework, HIDG further significantly reduces the total number of FEs for decomposition. Experiments on two sets of LSGO benchmark functions verified the effectiveness and efficiency of HIDG.

A global information based adaptive threshold for grouping large scale optimization problems

By taking the idea of divide-and-conquer, cooperative coevolution (CC) provides a powerful architecture for large scale global optimization (LSGO) problems, but its efficiency highly relies on the decomposition strategy. It has been shown that differential grouping (DG) performs well on decomposing LSGO problems by effectively detecting the interaction among decision variables. However, its decomposition accuracy highly depends on the threshold. To improve the decomposition accuracy of DG, a global information based adaptive threshold setting algorithm (GIAT) is proposed in this paper. On the one hand, by reducing the sensitivities of the indicator in DG to the roundoff error and the magnitude of contribution weight of subcomponent, we proposed a new indicator for two variables which is much more sensitive to their interaction. On the other hand, instead of setting the threshold only based on one pair of variables, the threshold is generated from the interaction information for all pair of variables. By conducting the experiments on two sets of LSGO benchmark functions, the correctness and robustness of this new indicator and GIAT were verified.

Anisotropic adaptive variance scaling for Gaussian estimation of distribution algorithm

Abstract Traditional Gaussian estimation of distribution algorithms (EDAs) are confronted with issues that the variable variances decrease fast and the main search direction tends to become perpendicular to the improvement direction of the fitness function, which reduces the search efficiency of Gaussian EDAs (GEDAs) and makes them subject to premature convergence. In this paper, a novel anisotropic adaptive variance scaling (AAVS) technique is proposed to improve the performance of traditional GEDAs and a new GEDA variant named AAVS-EDA is developed. The advantages of AAVS over the existing variance scaling strategies lie in its ability for tuning the variances and main search direction of GEDA simultaneously, which are achieved by anisotropically scaling the variances along different eigendirections based on corresponding landscape characteristics captured by a simple topology-based detection method. Besides, AAVS-EDA also adopts an auxiliary global monitor to ensure its convergence by shrinking all the variances if no improvement is achieved in a generation. The evaluation results on 30 benchmark functions of CEC2014 test suite demonstrate that AAVS-EDA possesses stronger global optimization efficiency than traditional GEDAs. The comparison with other state-of-the-art evolutionary algorithms also shows that AAVS-EDA is efficient and competitive.

Surrogate model assisted cooperative coevolution for large scale optimization

It has been shown that cooperative coevolution (CC) can effectively deal with large scale optimization problems (LSOPs) through a ‘divide-and-conquer’ strategy. However, its performance is severely restricted by the current context-vector-based sub-solution evaluation method, since this method needs to invoke the original high dimensional simulation model when evaluating each sub-solution, thus requiring many computation resources. To alleviate this issue, this study proposes a novel surrogate model assisted cooperative coevolution (SACC) framework. SACC constructs a surrogate model for each sub-problem and employs it to evaluate corresponding sub-solutions. The original simulation model is only adopted to reevaluate a small number of promising sub-solutions selected by surrogate models, and these really evaluated sub-solutions will in turn be employed to update surrogate models. By this means, the computation cost could be greatly reduced without significantly sacrificing evaluation quality. By taking the radial basis function (RBF) and the success-history based adaptive differential evolution (SHADE) as surrogate model and optimizer, respectively, this study further designs a concrete SACC algorithm named RBF-SHADE-SACC. RBF and SHADE have only been proved to be effective on small and medium scale problems. This study scales them up to LSOPs under the SACC framework, where they are tailored to a certain extent for adapting to the characteristics of LSOPs and SACC. Empirical studies on IEEE CEC 2010 benchmark functions demonstrate that SACC can significantly enhance the sub-solution evaluation efficiency, and even with much fewer computation resources, RBF-SHADE-SACC can find much better solutions than traditional CC algorithms.