Catálogo de publicaciones - libros

In thispaper, a approach for automatically generating fuzzy rules from sample patterns is presented. Then a self-adaptive fuzzy neural network is built based on the fuzzy partition which divides the input space with input and output information. The salient characteristics of the self-adaptive fuzzy neural networks are:1) structure identification and parameters estimation are performed automatically and simultaneously ;2)fuzzy rules can be recruited or deleted dynamically;3)parameters of rules can be obtained by evolutionary computation. Simulation results demonstrate that a compact and high performance fuzzy rule base can be constructed. Comprehensive comparisons with other approach show that the proposed approach is superior over other in terms of learning efficiency and performance.

The paper gives an adaptive particle swarm optimization algorithm with new random inertia weight (RIW-PSO). The new random inertia weight (RIW) is presented by simulated annealing idea to improve the global search ability of PSO and the one to solve the high dimensional and complex nonlinear optimization problems. The PSO with linearly decreasing inertia weight (LDWPSO) and RIW-PSO are tested with six benchmark functions. The experiments show that the convergent speed and accuracy of RIW-PSO is significantly superior to the one of LDW-PSO.

The recognition of human gestures in image sequences is an important and challenging problem that enables a host of human-computer interaction applications. This paper describes a gesture recognition algorithm based on the particle filters, namely CONDENSATION. The particle filter is more efficient than any other tracking algorithm because the tracking mechanism follows Bayesian estimation rule of conditional probability propagation. We used two models for the evaluation of particle filter and apply the MATLAB for the preprocessing of the image sequence. But we implement the particle filter using the C++ to get the high speed processing. In the experimental results, it is demonstrated that the proposed algorithm prove to be robust in the cluttered environment.

The paper presents a modified particle swarm optimization (PSO) for solving systems of equations problem (SEP). With the hope to improve the global performance of PSO, the modified method adopts traditional controller to control the search dynamics, such as PI or PID controller. Through the introduction of traditional controller, the modified PSO can feed back the search information to adjust the inertia weight adaptively, which in turn balances the global exploration and the local exploitation validly. Further more, the modified PSO takes advantage of a single neuron network as a learner to get appropriate parameters for the traditional controller. The modified PSO with controller has been applied to solve some systems of equations. The experimental results show the proposed method is efficient and robust for optimization.

In this paper, we focus on the optimization of fuzzy clustering. Particle Swarm Optimizations (PSO) is used for optimizing the algorithms. PSO is an algorithm which takes a cue from nature’s bird flock or fish school and is known to have superior ability in search and fast convergence. But it might be difficult to find global optimal solution of the fuzzy clustering when it comes to complex higher dimensions. So we optimize the fuzzy clustering using Predator Prey Particle Swarm Optimizations (PPPSO). The concept of PPPSO is that predators chase the center of prey’s swarm, and preys escape from predators, in order to avoid local optimal solutions and find global optimal solution efficiently.The performance of fuzzy c-means (FCM), particle swarm fuzzy clustering (PSFC) and predator prey particle swarm fuzzy clustering (PPPSFC) are compared. Through experiments, we show that the proposed algorithm has the best performance among them.

The paper presents an immune particle swarm optimization with diversity monitoring (DIPSO). In order to maintain appropriate diversity, DIPSO take advantage of immune operators to update the particles at the right moment through monitoring the swarm diversity of each generation. The modified algorithm can avoid the local optimum and has better search performance for multi-peak functions. Testing over the benchmark problems, the experimental results show the modified algorithm has better convergence performance than standard particle swarm optimization algorithm.

Particle Swarm Optimization algorithm (PSO) is a new evolutionary computation method, which has been successfully applied to many fields. However it also has problem of premature convergence and slow search speed. To deal with those problems we make some improvements on traditional PSO to make its search velocity quickly. Then we add some others algorithms and new ideas to PSO to construct a new PSO system (PSOS). Those algorithms and new ideas will be applied to one or several particles, which have their own specified duty or responsibility, and work in collaboration and communicate with others common particles in the PSOS. In the process of iterative computation, particles will keep updating their position according to specific circumstances until achieve their common purpose,that means finding out the global optimum solution quickly and exactly.

The particle swarm optimization algorithm is an algorithm to find optimal regions of complex spaces through the interaction of individuals. Convergence analysis and parameter selection in the particle swarm optimization algorithm have been discussed in [2] and [7]. In this paper, the particle swarm optimization algorithm is analyzed further by using standard results from the dynamic system theory. Thus, we derived graphical parameter guidelines from it. Finally, we analyze the convergence of the algorithm by some examples.

The training time for Support vector machine (SVM) depends largely on the size of the training set, which makes it impractical for large data sets. This paper presents a new method to reduce the size by combining two supplementary algorithms. The training data is partitioned into several pair-wise disjoint clusters by using -means clustering algorithm. Then, the representatives of these clusters can be edited by Gabriel graph algorithm, based on which we can approximately identify the support vectors and non-support vectors. After de-clustering the marginal boundary clusters represented by support vectors and deleting the internal clusters represented by non-support vectors, the number of training data can be significantly reduced, thereby speeding up the training process. The proposed method was tested on both the artificial data and real data. Experiment results show that replacing the training set with the edited set obtained from Gabriel graph algorithm and -means clustering technique as the training set, significantly reduces the training time for SVM, yet the classification accuracy remains nearly undegraded.

There is a large number of ordinary differential equations (ODEs) characterize the electrical behavior generated by ionic movements in human myocardial cell. In this paper, several approaches were investigated in order to improve the efficiency of solving the ODE systems for ten Tusscher et al.’s ionic model of human ventricular tissue. By using non-standard finite difference (NSFD) scheme, the stiffness of the ODEs system will be successfully reduced, so a larger step-size can be used. A popular multi-step method called backward differentiation formulation (BDF) was also incorporated into the computational model for testing the largest possible time steps. The results show that NSFD can be as much as 10 times more efficient than standard forward Euler in single cell model simulation while maintaining an acceptable level of accuracy. The investigation of BDF method shows that a large step size is not recommended for the single cell simulation application. All solvers were coupled to the partial differential equations for the complete simulation of heart tissue, and such computation scheme may be a good calculation technique in heart modeling and simulation.