In recent years the Probabilistic Neural Network (PPN) has been used in a large number of applications due to its simplicity and efficiency. PNN assigns the test data to the class with maximum likelihood compared with other classes. Likelihood of the test data to each training data is computed in the pattern layer through a kernel density estimation using a simple Bayesian rule. The kernel is usually a standard probability distribution function such as a Gaussian function. A spread parameter is used as a global parameter which determines the width of the kernel. The Bayesian rule in the pattern layer estimates the conditional probability of each class given an input vector without considering any probable local densities or heterogeneity in the training data. In this paper, an enhanced and generalized PNN (EPNN) is presented using local decision circles (LDCs) to overcome the aforementioned shortcoming and improve its robustness to noise in the data. Local decision circles enable EPNN to incorporate local information and non-homogeneity existing in the training population. The circle has a radius which limits the contribution of the local decision. In the conventional PNN the spread parameter can be optimized for maximum classification accuracy. In the proposed EPNN two parameters, the spread parameter and the radius of local decision circles, are optimized to maximize the performance of the model. Accuracy and robustness of EPNN are compared with PNN using three different benchmark

A new chaos-wavelet approach is presented for electroencephalogram (EEG)-based diagnosis of Alzheimer's disease (AD) employing a recently developed concept in graph theory, visibility graph (VG). The approach is based on the research ideology that nonlinear features may not reveal differences between AD and control group in the band-limited EEG, but may represent noticeable differences in certain sub-bands. Hence, complexity of EEGs is computed using the VGs of EEGs and EEG sub-bands produced by wavelet decomposition. Two methods are employed for computation of complexity of the VGs: one based on the power of scale-freeness of a graph structure and the other based on the maximum eigenvalue of the adjacency matrix of a graph. Analysis of variation is used for feature selection. Two classifiers are applied to the selected features to distinguish AD and control EEGs: a Radial Basis Function Neural Network (RBFNN) and a two-stage classifier consisting of Principal Component Analysis (PCA) and the RBFNN. After comprehensive statistical studies, effective classification features and mathematical markers were discovered. Finally, using the discovered features and a two-stage classifier (PCA-RBFNN), a high diagnostic accuracy of 97.7% was obtained.

No abstract

The superior colliculus is a layered structure important for body- and gaze-orienting responses. Its superficial layer is, next to the lateral geniculate nucleus, the second major target of retinal ganglion axons and is retinotopically organized. Here we show that in the mouse there is also a precise organization of orientation preference. In columns perpendicular to the tectal surface, neurons respond to the same visual location and prefer gratings of the same orientation. Calcium imaging and extracellular recording revealed that the preferred grating varies with retinotopic location, and is oriented parallel to the concentric circle around the centre of vision through the receptive field. This implies that not all orientations are equally represented across the visual field. This makes the superior colliculus different from visual cortex and unsuitable for translation-invariant object recognition and suggests that visual stimuli might have different behavioural consequences depending on their retinotopic location.