Arun D. Kulkarni

Fuzzy neural network models for classification

In this paper, we combine neural networks with fuzzy logic techniques. We propose a fuzzy-neural network model for pattern recognition. The model consists of three layers. The first layer is an input layer. The second layer maps input features to the corresponding fuzzy membership values, and the third layer implements the inference engine. The learning process consists of two phases. During the first phase weights between the last two layers are updated using the gradient descent procedure, and during the second phase membership functions are updated or tuned. As an illustration the model is used to classify samples from a multispectral satellite image, a data set representing fruits, and Iris data set.

Knowledge discovery from multispectral satellite images

A new approach to extract knowledge from multispectral images is suggested. We describe a method to extract and optimize classification rules using fuzzy neural networks (FNNs). The FNNs consist of two stages. The first stage represents a fuzzifier block, and the second stage represents the inference engine. After training, classification rules are extracted by backtracking along the weighted paths through the FNN. The extracted rules are then optimized by use of a fuzzy associate memory bank. We use the algorithm to extract classification rules from a multispectral image obtained with a Landsat Thematic Mapper sensor. The scene represents the Mississippi River bottomland area. In order to verify the rule extraction method, measures such as the overall accuracy, producers accuracy, users accuracy, kappa coefficient, and fidelity are used.

Fuzzy neural network models for supervised classification: Multispectral image analysis

Abstract It has been well established that neural networks provide a reasonable and powerful alternative to conventional classifiers. During the past few years there has been a large and energetic upswing in research efforts aimed at synthesizing fuzzy logic with neural networks. This combination of fuzzy logic and neural networks seems natural because two approaches generally attack the design of “intelligent” systems from quite different angles. Neural networks provide algorithms for learning, classification, and optimization whereas fuzzy logic deals with issues such as reasoning on a higher (semantic or linguistic) level. Consequently the two technologies complement each other. In this paper we propose two novel fuzzy‐neural network models for supervised learning. The first model consists of three layers, and the second model consists of four layers. In both models, the first two layers implement fuzzy membership functions and the remaining layers implement the inference engine. Both models use the gr...

Neural-Fuzzy Models for Multispectral Image Analysis

In this paper, we consider neural-fuzzy models for multispectral image analysis. We consider both supervised and unsupervised classification. The model for supervised classification consists of six layers. The first three layers map the input variables to fuzzy set membership functions. The last three layers implement the decision rules. The model learns decision rules using a supervised gradient descent procedure. The model for unsupervised classification consists of two layers. The algorithm is similar to competitive learning. However, here, for each input sample, membership functions of output categories are used to update weights. Input vectors are normalized, and Euclidean distance is used as the similarity measure. In this model if the input vector does not satisfy the “similarity criterion,” a new cluster is created; otherwise, the weights corresponding to the winner unit are updated using the fuzzy membership values of the output categories. We have developed software for these models. As an illustration, the models are used to analyze multispectral images.

Water Quality Retrieval from Landsat TM Imagery

Abstract In this paper, the utility of Landsat TM imagery for water quality studies in East Texas is investigated. Remote sensing has an important and effective role in water quality management. Remote sensing satellites measure the amount of solar radiation reflected by surface water and the reflectance of water depend upon the concentration and character of water quality parameters. Three water quality parameters namely the total suspended solids, chlorophyll-a, and turbidity are estimated in this study. In situ water quality parameter measurements from seven ground stations and the corresponding Landsat TM data were used to estimate the water quality parameters. Regression models are used to evaluate correlation between the water quality parameters and spectral reflectance values.

Artificial neural network models for texture classification via the radon transform

Texture is generally recognized as being fundamental to perception. A taxonomy of problems encountered within the context of texture analysis could be that of classification/discrimination, description, and segmentation. In this paper we suggest a novel artificial neural network (ANN) architecture for features extraction and texture recognition. There is evidence which suggests that the analysis of stimulus by visual system might involve a set of quasi-independent mechanisms called channels which could be conveniently characterized in the spatial frequency domain. In our model we use an FT feature space with angular and radial bins that characterize spatial domain filters to extract features. The extracted features are then used as input for the recognition stage. In order to evaluate the 2-D FT coefficients we use the Radon transform. The usage of the Radon transform simplifies the ANN model significantly. We suggest an electronic implementation of the ANN model for feature extraction, using a Connected Network Adaptive ProcessorS (CNAPS) chip designed by Adaptive Solutions Inc. We also develop software to simulate the ANN model with the Radon transform. We use a three stage back-propagation network as a classifier. We have used ten different texture patterns to test our ANN model.

Self organizing neural networks with a split/merge algorithm

In this paper we present a new learning algorithm for Artificial Neural Networks (ANN) using a split/merge technique. An ANN model with the new algorithm has been developed and tested on a PC. The model detects the similarity between the input patterns, and identifies the number of categories present in the input samples. The algorithm is similar to a competitive learning algorithm. Unlike the competitive learning algorithm, in this algorithm we use two types of weights: long term weights (LTWs) and short term weights (STWs). The network stability is provided by the LTWs, whereas the network plasticity is provided by STWs. As an illustration, the model is used to categorize pixels in a multispectral image. The categorization is based on the observed spectral signature at each pixel.

Neural nets for image restoration

No imaging system in practice is perfect, in fact the recorded images are always distorted or of finite resolution. An image recording system can be modeled by a Fredholm integral equation of the first kind. An inversion of the kernel representing the system, in the presence of noise, is an ill posed problem. The direct inversion often yields an unacceptable solution. In this paper, we suggest an Artificial Neural Network (ANN) architecture to solve ill posed problems in the presence of noise. We use two types of neuron like processing units: the units that use the weighted sum and the units that use the weighted product. The weights in the model are initialized using the eigen vectors of the kernel matrix that characterizes the recording system. We assume the solution to be a sample function of a wide sense stationary process with a known auto-correlation function. As an illustration, we consider two images that are degraded by motion blur.

Knowledge Extraction from Survey Data Using Neural Networks

Abstract Surveys are an important tool for researchers. It is increasingly important to develop powerful means for analyzing such data and to extract knowledge that could help in decision-making. Survey attributes are typically discrete data measured on a Likert scale. The process of classification becomes complex if the number of survey attributes is large. Another major issue in Likert-Scale data is the uniqueness of tuples. A large number of unique tuples may result in a large number of patterns. The main focus of this paper is to propose an efficient knowledge extraction method that can extract knowledge in terms of rules. The proposed method consists of two phases. In the first phase, the network is trained and pruned. In the second phase, the decision tree is applied to extract rules from the trained network. Extracted rules are optimized to obtain a comprehensive and concise set of rules. In order to verify the effectiveness of the proposed method, it is applied to two sets of Likert scale survey data, and results show that the proposed method produces rule sets that are comparable with other knowledge extraction techniques in terms of the number of rules and accuracy.

A testbed for sensory PDP models

The Parallel Distributed Processing model of cognition has provided a number of new algorithms for Artificial Intelligence. It has been especially successful in the areas of computer vision and pattern recognition. This paper describes a general PDP model that can be used for sensory systems including vision, olfactory and language. The model has a number of parameters and we are using it as a testbed to determine the characteristics of these parameters.