A gene selection algorithm for microarray cancer classification using an improved particle swarm optimization

BackgroundGene selection is one of the critical steps in the course of the classification of microarray data. Since particle swarm optimization has no complicated evolutionary operators and fewer parameters need to be adjusted, it has been used increasingly as an effective technique for gene selection. Since particle swarm optimization is apt to converge to local minima which lead to premature convergence, some particle swarm optimization based gene selection methods may select non-optimal genes with high probability. To select predictive genes with low redundancy as well as not filtering out key genes is still a challenge.ResultsTo obtain predictive genes with lower redundancy as well as overcome the deficiencies of traditional particle swarm optimization based gene selection methods, a hybrid gene selection method based on gene scoring strategy and improved particle swarm optimization is proposed in this paper. To select the genes highly related to out samples’ classes, a gene scoring strategy based on randomization and extreme learning machine is proposed to filter much irrelevant genes. With the third-level gene pool established by multiple filter strategy, an improved particle swarm optimization is proposed to perform gene selection. In the improved particle swarm optimization, to decrease the likelihood of the premature of the swarm the Metropolis criterion of simulated annealing algorithm is introduced to update the particles, and the half of the swarm are reinitialized when the swarm is trapped into local minima.ConclusionsCombining the gene scoring strategy with the improved particle swarm optimization, the new method could select functional gene subsets which are significantly sensitive to the samples’ classes. With the few discriminative genes selected by the proposed method, extreme learning machine and support vector machine classifiers achieve much high prediction accuracy on several public microarray data, which in turn verifies the efficiency and effectiveness of the proposed gene selection method.

Efficient multiprocessor scheduling is essentially the problem of allocating a set of computational jobs to a set of processors to minimize the overall execution time. The main issue is how jobs are partitioned in which total finishing time and waiting time is minimized. Minimization of these two criteria simultaneously, is a multi objective optimization problem. There are many variations of this problem, most of which are NP-hard problem, so we must rely on heuristics to solve the problem instances. Many heuristic-based approaches have been applied to finding schedules that minimize the execution time of computing tasks on parallel processors. Particle swarm optimization (PSO) is currently employed in several optimization and search problems due to its ease and ability to find solutions successfully. A variant of PSO, called as improved particle swarm optimization (ImPSO) has been developed in this paper and is hybridized with the ant colony optimization (ACO) to achieve better solutions. The proposed hybrid algorithm effectively exploits the capabilities of distributed and parallel computing of swarm intelligence approaches. In addition hybrid algorithm using improved particle swarm optimization (ImPSO) with artificial immune system (AIS) is also implemented for the same set of problems to compare with the proposed hybrid algorithm (ImPSO with ACO). It was observed that the proposed hybrid approach (Improved PSO with ACO) gives better results in experiments and reduces finishing and waiting time simultaneously. Key words: Particle swarm optimization (PSO), improved particle swarm optimization (ImPSO), ant colony optimization (ACO), job scheduling, finishing time, waiting time.

Aiming at the control variables of reactive power optimization are discrete, and some parameters in the standard particle swarm optimization (PSO) algorithm need to be predefined by test, so the algorithm's practicability is restricted. For these reasons, an adaptive particle swarm optimization (APSO) algorithm is proposed by the authors. APSO introduces the self-adaptive tuning strategy and boundary constraint conditions can find the global optimal solution and solve the discrete variables. The reactive power optimization results of the standard IEEE-30-bus power system show that APSO is efficient than standard PSO. The global convergence accuracy and convergence stability is obviously improved compared with that of PSO.

An improved particle swarm optimization (PSO) method applied to the design of a new wideband log-periodic antenna (LPA) geometry is introduced. This new PSO variant, called PSO with velocity mutation (PSOvm), induces mutation on the velocities of those particles that cannot improve their position. The proposed LPA consists of wire dipoles with lengths and distances varied according to an exponential rule, which is defined by two specific parameters called length factor and spacing factor. The LPA is optimized for operation in 790-6000 MHz frequency range, in order to cover the most usual wireless services in practice, and also to provide in this range the highest possible forward gain, gain flatness below 2 dB, secondary lobe level below -20 dB with respect to the main lobe peak, and standing wave ratio below 2. To demonstrate its superiority in terms of performance, PSOvm is compared with well-known optimization methods. The comparison is performed by applying all the methods on several test functions and also on the LPA optimization problem defined by the above-mentioned requirements. Furthermore, the radiation characteristics of the PSOvm-based LPA give prominence to the effectiveness of the proposed exponential geometry compared to the traditional Carrel's geometry.

In this article we describe a novel Adaptive particle swarm optimization (APSO) algorithm based on population diversity information. It is presented to solve the precocious convergence problem of particle swarm optimization algorithm. The APSO algorithm uses the information of the population diversity to adjust nonlinearly inertia weight. Velocity mutation factor and position interchange factor are both introduced and the global performance is clearly improved. The APSO algorithm is applied to optimization of parameters in the optimal model based on support vector machine (SVM). SVM is a popular classification method with many diverse applications. A novel Adaptive particle swarm optimization (APSO) based approach for parameter determination and feature selection of the SVM, termed APSO+SVM is developed. The illustrating example shows that the classification accuracy of APSO+SVM is higher than other traditional methods of classification, so using APSO+SVM method to classify is feasible and effective.

In this paper, we present an improved particle swarm optimization (PSO) algoritlim to solve constrained optimization problems. The proposed approach, called MPSO, employs a novel mutation operator to enhance the global search ability of PSO. In order to deal with constrains, MPSO uses mean violations mechanism and boundaries search. Simulation results on five famous benchmark problems show that MPSO achieves better results than standard PSO and another variant of PSO.

Network reconfiguration for loss reduction in a distribution system is a very important way to save the electrical energy. This article proposes a hybrid evolutionary algorithm to solve the distribution feeder reconfiguration problem. The algorithm combines a fuzzy adaptive particle swarm optimization with a differential evolution algorithm, called the fuzzy adaptive particle swarm optimization–differential evolution. The fuzzy adaptive particle swarm optimization includes two parts. The first part is fuzzy adaptive binary particle swarm optimization, which determines the status of tie switches (open or closed), and the second part is fuzzy adaptive discrete particle swarm optimization, which determines the sectionalizing switch number. In the proposed algorithm, due to the results of binary particle swarm optimization and discrete particle swarm optimization algorithms that highly depend on the values of their parameters (such as the inertia weight and learning factors), a fuzzy system is employed to adaptively adjust the parameters during the search process. The differential evolution algorithm is combined with the fuzzy adaptive particle swarm optimization algorithm to improve its performance. The proposed algorithm is tested on two distribution test feeders. The results of simulation show that the proposed method is very powerful and is guaranteed to obtain the global optimization.

This paper compares the performance of two variants of the Particle Swarm Optimization (PSO) algorithm; PSO with constriction factor (PSO), and mutative PSO (MPSO) in optimizing Mel Frequency Cepstrum Coefficients (MFCC) parameters. The parameters were used to extract an optimal feature set for classifying healthy and hypothyroid infant cry using Multi-Layer Perceptrons (MLP). Specifically, the PSO variants optimize the number of filter banks and number of cepstrum coefficients in MFCC. Based on the values chosen by both PSO variants, the extracted features were then fed to a MLP classifier, which was trained to discriminate between the healthy and hypothyroid infant cry. Comparisons between the performance of PSO variants showed that MPSO managed to improve the convergence rate by 2.67% compared to PSO.

Experimental study of a new hybrid PSO with mutation for economic dispatch with non-smooth cost function

Particle Swarm Optimization (PSO) is a biologically inspired computational search and optimization method developed in 1995 by Eberhart and Kennedy based on the social behaviors of birds flocking or fish schooling.A number of basic variations have been developed due to improve speed of convergence and quality of solution found by the PSO.On the other hand, basic PSO is more appropriate to process static, simple optimization problem.Modification PSO is developed for solving the basic PSO problem.The observation and review 46 related studies in the period between 2002 and 2010 focusing on function of PSO, advantages and disadvantages of PSO, the basic variant of PSO, Modification of PSO and applications that have implemented using PSO.The application can show which one the modified or variant PSO that haven't been made and which one the modified or variant PSO that will be developed.

Binocular vision technology is widely used in daily life, and binocular camera calibration is of great significance. Aiming at the problem that the training accuracy of the current camera calibration method based on neural network is not high enough, this paper proposes a binocular camera calibration method based on the improved particle swarm optimization BP neural network (APSO-BPNN), and introduces the adaptive particle swarm optimization algorithm to optimize the BP neural network. A binocular camera calibration model is constructed, and real data are used to verify the effectiveness and feasibility of the method. The experimental results show that the dual-target calibration method based on APSO-BPNN has higher calibration accuracy than the traditional neural network calibration method.

A dynamic adaptive particle swarm optimization and genetic algorithm is presented to solve constrained engineering optimization problems. A dynamic adaptive inertia factor is introduced in the basic particle swarm optimization algorithm to balance the convergence rate and global optima search ability by adaptively adjusting searching velocity during search process. Genetic algorithm–related operators including a selection operator with time-varying selection probability, crossover operator, and n-point random mutation operator are incorporated in the particle swarm optimization algorithm to further exploit optimal solutions generated by the particle swarm optimization algorithm. These operators are used to diversify the swarm and prevent premature convergence. Tests on nine constrained mechanical engineering design optimization problems with different kinds of objective functions, constraints, and design variables in nature demonstrate the superiority of the dynamic adaptive particle swarm optimization and genetic algorithm against several other meta-heuristic algorithms in terms of solution quality, robustness, and convergence rate in most cases.

An adaptive dynamic multi-swarm particle swarm optimization with stagnation detection and spatial exclusion for solving continuous optimization problems

Several studies exist in the literature that utilized metaheuristics and nature-inspired algorithms to solve engineering problems. Particle Swarm Optimization (PSO) is a well-known nature-inspired algorithm that has been used for different optimization problems due to its simplicity and ability to find near-optimal solutions. However, PSO suffers from a problem of balancing between the global search and local search abilities when applied to engineering problems. Recently, a new variant of PSO based on a novel multi-swarm architecture called Multi-swarm Particle Swarm Optimization (MPSO) was proposed. The proposed MPSO was evaluated on solving normal and large-scale problems. This study evaluated the possibility of using MPSO to simulate Welded Beam Design (WDB) problem which is a mechanical engineering problem. Several simulations were performed using the proposed approach and from the results, MPSO model was observed to simulate WBD problem with better optimal solution compared to the standalone PSO. The outcome of this study further confirmed the robustness of the MPSO over the other known metaheuristics. Generally, MPSO achieved an excellent optimization performance with a fast convergence learning process.

Due to the disadvantages of the canonical Particle Swarm Optimization (PSO), a new adaptive particle swarm optimization algorithm based on Sigmoid inertia weight was proposed. On the one hand, the premature computing formula of the particles was introduced so as to conduct the evolution process and enhance the convergent speed. On the other hand, the decision threshold of particles' focusing degree was employed in order to make the PSO adaptively adopt the inertia weight between linear decrease and nonlinear decrease based on Sigmoid function, which could effectively prevent the PSO from plunging into local minimum. The experimental results show that the proposed algorithm is feasible and effective.