This paper proposes a novel regularization approach for Extreme Learning Machines. Regularization is performed using a priori spacial information expressed by an affinity matrix. We show that the use of this type of a priori information is similar to perform Tikhonov regularization. Furthermore, if a parameter free affinity matrix is used, like the cosine similarity matrix, regularization is pe...

We present molecular dynamics simulations with empirical potentials to study the type of defects produced when irradiating graphene with low energy carbon ions (100 eV and 200 eV) and different dose rates. Simulations show the formation of very stable structures such as dimers, single chains of carbons and double chains of carbons. These structures are similar to those described in the literatu...

In this paper we propose a supervised version of the Isomap algorithm by incorporating class label information into a dissimilarity matrix in a financial analysis setting. On the credible assumption that corporates financial status lie on a low dimensional manifold, nonlinear dimensionality reduction based on manifold learning techniques has strong potential for bankruptcy analysis in financial...

During the last two years several methods have been proposed for handling nonlinear constraints by genetic algorithms for numerical optimization problems; most of them were based on penalty functions. However, the performance of these methods is highly problem-dependent; moreover, many methods require additional tunning of several parameters. In this paper we present a new optimization system (...

Exact solutions of conformal turbulence restricted on a upper half plane are obtained. We show that the inertial range of homogeneous and isotropic turbulence with constant enstrophy flux develops in a distant region from the boundary. Thus in the presence of an anisotropic boundary, these exact solutions of turbulence generalize Kolmogorov’s solution consistently and differ from the Polyakov’s...

In this paper, a PID fuzzy controller structure is presented. To be successful in implementing fuzzy PID controllers in large applications we must also considered their operating principles. Fuzzy logic controllers are based on empirical control rules. A study case for the validation of the proposed tunning mechanism is presented and analyzed with control application for a synchronous generator...

During the last two years several methods have been proposed for handling nonlinear constraints b y genetic algorithms for numerical optimization problems; most of them were based on penalty functions. However, the performance of these methods is highly problem-dependent; moreover, many methods require additional tunning of several parameters. In this paper we present a new optimization system ...

This paper presents an approach to the design of self-tuning adaptive power system stabilizer which is based on artificial neural network. Result shows that ANN based power system stabilizer can provide good damping for both local and inter area modes of oscillations. An ANN is used for self-tuning the different parameters of PSS like stabilizing gain Kstab and time constant (T1) for Lead PSS i...