Catálogo de publicaciones - libros

This paper studies the KRW-metric stability of Hopfield stochastic neural networks dx ( t ) = [ −  Ax ( t ) +  Bg ( x ( t ))] dt  +  σ ( x ( t )) dw ( t ). From the probabilistic point of view, the KRW-metric stability have an intrinsic property which makes them more suitable for certain applications. But up to now, the results about KRW-metric stability are very little. In this paper, by the use of the contraction of diffusion semigroup with to Lipschitz constant and coupling technique we obtain some sufficient conditions for stability and exponential stability. The presented results do not require constructing Lyapunov functions what we need is only to compute some estimates which depend only on the coefficients of equations.

Using the classical Parzen window (PW) estimate as the target function, the sparse kernel density estimator is constructed in a forward constrained regression manner. The leave-one-out (LOO) test score is used for kernel selection. The jackknife parameter estimator subject to positivity constraint check is used for the parameter estimation of a single parameter at each forward step. As such the proposed approach is simple to implement and the associated computational cost is very low. An illustrative example is employed to demonstrate that the proposed approach is effective in constructing sparse kernel density estimators with comparable accuracy to that of the classical Parzen window estimate.

Active Appearance Model (AAM) is a generic model for the certain visual phenomena, and one of the powerful modeling techniques in the image processing area. AAM has been initially developed for the face modeling. Recently, it is shown that it may be useful for other area. In this paper, we first present how to use AAM in facial expression recognition and then extend it to the left ventricle segmentation problem. For the former, we establish face model using Cohn-Kanade facial expression database, whereas for the latter, SPEC images will be used for modeling the left ventricle. We show that the facial expression model can continuously track human facial expressions, and the left ventricle model can segment the inside and outside of the left ventricle. Our examples can be extended to other medical applications.