This paper presents an algorithm for the design of a computer aided diagnosis system to detect, quantify and classify the lesions of non-proliferative diabetic retinopathy as well as dry age related macular degeneration from the fundus retina images. Symptoms of non-proliferative diabetic retinopathy in images consist of bright lesions like hard exudates, cotton wool spots and dark lesions like microaneurysms, hemorrhages. Dry age related macular degeneration is manifested as a bright lesion called drusen. The proposed system consists of two parts: image processing, where preprocessed gray scale images are segmented to extract candidate lesions using a combination of Gaussian filtering and multilevel thresholding followed by classification of the different lesions in non-proliferative diabetic retinopathy and age related macular degeneration using perceptron, support vector machine and naive Bayes classifier. From the comparative performance analysis of the classification techniques, it is observed that comparable results are obtained from single layer perceptron and support vector machine and they both outperform naive Bayes classifier. The classification accuracy of support vector machine classifier for dark lesion class is 97.13% and the classification accuracy of single layer perceptron for bright lesion class is 95.13% with optimal feature set.
In this paper, we report on experiments in which we used neural
networks for statistical anomaly intrusion detection systems. The five types of neural networks that we studied were: Perceptron; Backpropagation; Perceptron-Backpropagation-Hybrid; Fuzzy ARTMAP; and Radial-Based Function. We collected four separate data sets from different simulation scenarios, and these data sets were used to test various neural networks with different hidden neurons. Our results showed that the classification capabilities of BP and PBH outperform those of other neural networks.
There are two basic types of artificial neural networks: Multi-Layer Perceptron (MLP) and Radial Basis Function network (RBF). The first type (MLP) consists of one type of neuron, which can be decomposed into a linear and sigmoid part. The second type (RBF) consists of two types of neurons: radial and linear ones. The radial basis function is analyzed and then used for decomposition of RBF network. The resulting Perceptron Radial Basis Function Network (PRBF) consists of two types of neurons: linear and extended sigmoid ones. Any RBF network can be directly converted to a four-layer PRBF network while any MLP network with three layers can be approximated by a five-layer PRBF network. The new PRBF network is then a generalization of MLP and RBF network abilities. Learning strategies are also discussed. The new type of PRBF network and its learning via repeated local optimization is demonstrated on a numerical example together with RBF and MLP for comparison. This paper is organized as follows: Basic properties of MLP and RBF neurons are summarized in the first two chapters. The third chapter includes novel relationship between sigmoidal and radial functions, which is useful for RBF decomposition and generalization. Description of new PRBF network, together with its properties, is subject of the fourth chapter. Numerical experiments with a PRBF and their requests are given in the last chapters.
Generalization phenornena which také plače in two different assembly neural networks are considered in the paper. Either of these two assembly networks is artificially partitioned into several subiietworks according to the number of classes that the network has to recognize. Hebb’s cissernblies are formed in the networks. One of the assembly networks is with binary connections, the other is with analog ones. Recognition abilities of the networks are compared on the task of handwritten character recognition. The third neural network of a perceptron type is considered in the paper for comparison with the previous ones. This latter network works according to the nearest-neighbor method. Computer simulation of all three neural networks was performed. Experirnents showed that the assembly network with binary connections has approximately the same recognition accuracy as the network realizing the nearest-neighbor technique.