Sharif Digital Repository

The problem of clustering with side information has received much recent attention and metric learning has been considered as a powerful approach to this problem. Until now, various metric learning methods have been proposed for semi-supervised clustering. Although some of the existing methods can use both positive (must-link) and negative (cannot-link) constraints, they are usually limited to learning a linear transformation (i.e., finding a global Mahalanobis metric). In this paper, we propose a framework for learning linear and non-linear transformations efficiently. We use both positive and negative constraints and also the intrinsic topological structure of data. We formulate our metric... 

We present a new filter trust-region approach for solving unconstrained nonlinear optimization problems making use of the filter technique introduced by Fletcher and Leyffer to generate non-monotone iterations. We also use the concept of a multidimensional filter used by Gould et al. (SIAM J. Optim. 15(1):17-38, 2004) and introduce a new filter criterion showing good properties. Moreover, we introduce a new technique for reducing the size of the filter. For the algorithm, we present two different convergence analyses. First, we show that at least one of the limit points of the sequence of the iterates is first-order critical. Second, we prove the stronger property that all the limit points... 

This paper presents a novel opposition-based self-adaptive modified gravitational search algorithm (OSAMGSA) for optimal reactive power dispatch and voltage control in power-system operation. The problem is formulated as a mixed integer, nonlinear optimization problem, which has both continuous and discrete control variables. In order to achieve the optimal value of loss, voltage deviation, and voltage stability index, it is necessary to find the optimal value of control variables such as the tap positions of tap changing transformers, generator voltages, and compensation capacitor. Therefore, this complicated problem needs to be solved by an accurate optimization algorithm. This paper... 

We consider nonlinear optimization problems constrained by a system of fuzzy relation equations. The solution set of the fuzzy relation equations being nonconvex, in general, conventional nonlinear programming methods are not practical. Here, we propose a genetic algorithm with max-product composition to obtain a near optimal solution for convex or nonconvex solution set. Test problems are constructed to evaluate the performance of the proposed algorithm showing alternative solutions obtained by our proposed model  

In this paper, the order distribution concept in the frequency domain identification has been extended to include fractional order systems having poles and zeros simultaneously. The existing nonlinear optimization problem appeared when both poles and zeros, is are changed to a quadratic problem that can be solved using least squares algorithms. To collect the required data, system is excited by a multi sine input signal with appropriately selected frequencies. Then a nonparametric identification in frequency domain is accomplished to calculate the empirical transfer function estimate (ETFE). This estimate is then used to implement the frequency domain identification on all defined members of... 

Nature, as a rich source of solutions, can be an inspirational guide to answer scientific expectations. Seed dispersal mechanism as one of the most common reproduction method among the plants is a unique technique with millions of years of evolutionary history. In this paper, inspired by plants survival, a novel method of optimization is presented, which is called Fertile Field Algorithm. One of the main challenges of stochastic optimization methods is related to the efficiency of the searching process for finding the global optimal solution. Seeding procedure is the most common reproduction method among all the plants. In the proposed method, the searching process is carried out through a... 

This research study presents an investigation on the behavior of a Pulsating Heat Pipe Flat-Plate Solar Collector (PHPFPSC) by artificial neural network method and an optimization of the parameters of the collector by genetic algorithm. In this study, several experiments were performed to study the effects of various evaporator lengths, filling ratios, inclination angles, solar radiation, and input chilled water temperature between 9:00 A.M. to 5:00 P.M., and the output temperature of the water tank, which was the output of the system, was also measured. According to the input and output information, multilayer perceptron neural network was trained and used to predict the behavior of the...