An alternating projected gradient algorithm for nonnegative matrix factorization

Due to the extensive applications of nonnegative matrix factorizations (NMFs) of nonnegative matrices, such as in image processing, text mining, spectral data analysis, speech processing, etc., algorithms for NMF have been studied for years. In this paper, we propose a new algorithm for NMF, which is based on an alternating projected gradient (APG) approach. In particular, no zero entries appear in denominators in our algorithm which implies no breakdown occurs, and even if some zero entries appear in numerators new updates can always be improved in our algorithm. It is shown that the effect of our algorithm is better than that of Lee and Seung’s algorithm when we do numerical experiments on two known facial databases and one iris database.     

Using genetic algorithms for the coordinated scheduling problem of a batching machine and two-stage transportation

This paper considers a coordinated scheduling problem. For the first-stage transportation there is a crane available to transport the product from the warehouse to a batching machine. For the second-stage transportation there is a vehicle available to deliver the completed jobs from the machine shop floor to the customer. The coordinated scheduling problem of production and transportation deals with sequencing the transportation of the jobs and combining them into batches to be processed. The problem of minimizing the sum of the makespan and the total setup cost was proven by Tang and Gong [1] to be strongly NP-hard. This paper proposes two genetic algorithm (GA) approaches for this scheduling problem, with different result representations. The experimental results demonstrate that a regular GA and a modified GA (MGA) can find near-optimal solutions within an acceptable amount of computational time. Among the two proposed metaheuristic approaches, the MGA is superior to the GA both in terms of computing time and the quality of the solution.     

Quantum-inspired space search algorithm (QSSA) for global numerical optimization

This work presents the evolutionary quantum-inspired space search algorithm (QSSA) for solving numerical optimization problems. In the proposed algorithm, the feasible solution space is decomposed into regions in terms of quantum representation. As the search progresses from one generation to the next, the quantum bits evolve gradually to increase the probability of selecting the regions that render good fitness values. Through the inherent probabilistic mechanism, the QSSA initially behaves as a global search algorithm and gradually evolves into a local search algorithm, yielding a good balance between exploration and exploitation. To prevent a premature convergence and to speed up the overall search speed, an overlapping strategy is also proposed. The QSSA is applied to a series of numerical optimization problems. The experiments show that the results obtained by the QSSA are quite competitive compared to those obtained using state-of-the-art IPOP-CMA-ES and QEA.     

Discretization algorithm for fractional order integral by Haar wavelet approximation

A discretization algorithm is proposed by Haar wavelet approximation theory for the fractional order integral. In this paper, the integration time is divided into two parts, one presents the effect of the past sampled data, calculated by the iterative method, and the other presents the effect of the recent sampled data at a fixed time interval, calculated by the Haar wavelet. This method can reduce the amount of the stored data effectively and be applied to the design of discrete-time fractional order PID controllers. Finally, several numerical examples and simulation results are given to illustrate the validity of this discretization algorithm.     

Convergence theorems for mixed equilibrium problems, variational inequality problem and uniformly quasi-?-asymptotically nonexpansive mappings

We introduce a modified block hybrid projection algorithm for finding a common element of the set of common fixed points of an infinite family of closed and uniformly quasi-?-asymptotically nonexpansive mappings, the set of the variational inequality for an α-inverse-strongly monotone operator, the set of solutions of the mixed equilibrium problems. Then, we obtain strong convergence theorems for the sequences generated by this process in a 2-uniformly convex and uniformly smooth Banach space. Our results extend and improve ones from several earlier works.     

Numerical treatment of solving singular integral equations by using Sinc approximations

In this paper we propose numerical treatment for singular integral equations. The methods are developed by means of the Sinc approximation with smoothing transformations. Such approximation is an effective technique against the singularities of the equations, and achieves exponential convergence. Therefore the methods improve conventional results where only polynomial convergence have been reported. The resulting algebraic system is solved by least squares approximation and leap frog algorithm. Estimation of errors of the approximate solution is presented. Some experimental tests are presented to show the efficient of the proposed methods.     

Heteroscedastic normal-exponential mixture models: Bayesian and classical approaches

In this paper, we introduce a Bayesian analysis for mixture of distributions belonging to the exponential family. As a special case we consider a mixture of normal exponential distributions including joint modeling of the mean and variance. We also consider joint modeling of the mean and variance heterogeneity. Markov Chain Monte Carlo (MCMC) methods are used to obtain the posterior summaries of interest. We also introduce and apply an EM algorithm, where the maximization is obtained applying the Fisher scoring algorithm. Finally, we also include analysis of real data sets to illustrate the proposed methodology.     

An improved artificial bee colony algorithm for directing orbits of chaotic systems

Artificial bee colony algorithm (ABC) is a relatively new optimization technique which has been shown to be competitive to other population-based algorithms. However, there is still an insufficiency in ABC regarding its solution search equation, which is good at exploration but poor at exploitation. To address this concerning issue, we propose an improved ABC (IABC) by using a modified search strategy to generate a new food source in order that the exploration and exploitation can be well balanced and satisfactory optimization performances can be achieved. In addition, to enhance the global convergence, when producing the initial population, both opposition-based learning method and chaotic maps are employed. In this paper, the proposed algorithm is applied to control and synchronization of discrete chaotic systems which can be formulated as both multimodal numerical optimization problems with high dimension. Numerical simulation and comparisons with some typical existing algorithms demonstrate the effectiveness and robustness of the proposed approach.     

Determining Type II sensitivity ranges of the fractional assignment problem

This paper proposes iterative labeling algorithms to determine the Type II sensitivity ranges of the fractional assignment problem. Unlike the traditional sensitivity range which keeps the current optimal basis remaining optimal, the Type II sensitivity range is the range that keeps the current optimal assignment remaining optimal. Focusing only on the non-degenerate basic variables makes the Type II sensitivity range more practical. Three cases of perturbation, each with two kinds, are discussed. An example is presented to demonstrate the proposed algorithms.     

MB-GNG: Addressing drawbacks in multi-objective optimization estimation of distribution algorithms

We examine the model-building issue related to multi-objective estimation of distribution algorithms (MOEDAs) and show that some of their, as yet overlooked, characteristics render most current MOEDAs unviable when addressing optimization problems with many objectives. We propose a novel model-building growing neural gas (MB-GNG) network that is specially devised for properly dealing with that issue and therefore yields a better performance. Experiments are conducted in order to show from an empirical point of view the advantages of the new algorithm.