Ch. Sanjeev Kumar Dash

DE+RBFNs based classification: A special attention to removal of inconsistency and irrelevant features

A novel approach for the classification of both balanced and imbalanced dataset is developed in this paper by integrating the best attributes of radial basis function networks and differential evolution. In addition, a special attention is given to handle the problem of inconsistency and removal of irrelevant features. Removing data inconsistency and inputting optimal and relevant set of features to a radial basis function network may greatly enhance the network efficiency (in terms of accuracy), at the same time compact its size. We use Bayesian statistics for making the dataset consistent, information gain theory (a kind of filter approach) for reducing the features, and differential evolution for tuning center, spread and bias of radial basis function networks. The proposed approach is validated with a few benchmarked highly skewed and balanced dataset retrieved from University of California, Irvine (UCI) repository. Our experimental result demonstrates promising classification accuracy, when data inconsistency and feature selection are considered to design this classifier.

Differential Evolution-Based Optimization of Kernel Parameters in Radial Basis Function Networks for Classification

In this paper a two phases learning algorithm with a modified kernel for radial basis function neural networks is proposed for classification. In phase one a new meta-heuristic approach differential evolution is used to reveal the parameters of the modified kernel. The second phase focuses on optimization of weights for learning the networks. Further, a predefined set of basis functions is taken for empirical analysis of which basis function is better for which kind of domain. The simulation result shows that the proposed learning mechanism is evidently producing better classification accuracy vis-A -vis radial basis function neural networks (RBFNs) and genetic algorithm-radial basis function (GA-RBF) neural networks.

Radial basis function neural networks: a topical state-of-the-artsurvey

Abstract Radial basis function networks (RBFNs) have gained widespread appeal amongst researchers and have shown good performance in a variety of application domains. They have potential for hybridization and demonstrate some interesting emergent behaviors. This paper aims to offer a compendious and sensible survey on RBF networks. The advantages they offer, such as fast training and global approximation capability with local responses, are attracting many researchers to use them in diversified fields. The overall algorithmic development of RBF networks by giving special focus on their learning methods, novel kernels, and fine tuning of kernel parameters have been discussed. In addition, we have considered the recent research work on optimization of multi-criterions in RBF networks and a range of indicative application areas along with some open source RBFN tools.

Neural Networks Training Based on Differential Evolution in Radial Basis Function Networks for Classification of Web Logs

With the fastest growth of World Wide Web it is quite difficult to track and understand users’ need for the owners of a website. Hence, an intelligent analyzer is proposed to find out the browsing patterns of a user. Moreover the pattern, which is revealed from this deluge of web access logs must be interesting, useful, and understandable. In this paper, a two phases learning algorithm with a modified kernel for radial basis function neural networks is proposed to classify the web pages on time of access and region of access. In phase one a meta-heuristic approach known as differential evolution is used to reveal the parameters of the modified kernel. The second phase focus on optimization of weights for learning the networks. The simulation result shows that the proposed learning mechanism is evidently producing better classification accuracy vis-a-vis radial basis function neural networks.

Design of self-adaptive and equilibrium differential evolution optimized radial basis function neural network classifier for imputed database

A medoid based imputation is newly developed.A novel self-adaptive and equilibrium DE algorithm is designed for optimizing RBFNs.SAEDE-RBFN is applied on Knn, mean, medoid based imputed database for classification. The occurrence of missing values is not uncommon in real life databases like industrial, medical, and life science. The imputation of these values has been realized through the mean/mode of known values (for a quantitative/qualitative attribute) or nearest neighbors. Mean based imputation considerably underestimates the population variance and tends to weaken the attribute relationships. Similarly, the nearest neighbor approach uses only information of the nearest neighbors and leaving other observations aside. Hence to overcome the shortcomings of these methods, we have introduced a method known as medoid based imputation to impute missing values. Further, to achieve better performance, we have devised a novel classifier for imputed datasets, by using the self-adaptive control parameters of differential evolution (DE) with equilibrium of exploitation and exploration optimized radial basis function neural networks (RBFNs). By newly associating a weight parameter with target vector during mutation, we maintain equilibrium on the exploration and exploitation mechanism of DE. The self-adaptive equilibrium DE (SAEDE) is used to explore and exploit the suitable kernel parameters of RBFNs along with bias and then used for classifying unknown samples. The performance of the proposed classifier named as SAEDE-RBFN has been extensively evaluated on seven datasets retrieved from University of California, Irvine (UCI) and KEEL machine learning repositories after imputation by mean, nearest neighbor, and proposed method. The average performance of classifiers has been listed based on the imputation by K-nearest neighbor (Knn = 1, Knn = 3, Knn = 5, and Knn = 7), mean, and medoid, respectively. Outcome of the experimental study shows that the performance of SAEDE-RBFN on medoid based imputed dataset is relatively better than DE-RBFN.

An Empirical Analysis of Evolved Radial Basis Function Networks and Support Vector Machines with Mixture of Kernels

Classification is one of the most fundamental and formidable tasks in many domains including biomedical. In biomedical domain, the distributions of data in most of the datasets into predefined number of classes is significantly different (i.e., the classes are distributed unevenly). Many mathematical, statistical, and machine learning approaches have been developed for classification of biomedical datasets with a varying degree of success. This paper attempts to analyze the empirical performance of two forefront machine learning algorithms particularly designed for classification problem by adding some novelty to address the problem of imbalanced dataset. The evolved radial basis function network with novel kernel and support vector machine with mixture of kernels are suitably designed for the purpose of classification of imbalanced dataset. The experimental outcome shows that both algorithms are promising compared to simple radial basis function neural networks and support vector machine, respectively. However, on an average, support vector machine with mixture kernels is better than evolved radial basis function neural networks.

Towards Crafting a Smooth and Accurate Functional Link Artificial Neural Networks Based on Differential Evolution and Feature Selection for Noisy Database

AbstractThis work presents an accurate and smooth functional link artificial neural network (FLANN) for classification of noisy database. The accuracy and smoothness of the network is taken birth by suitably tuning the parameters of FLANN using differential evolution and filter based feature selection. We use Qclean algorithm for identification of noise, information gain theory for filtering irrelevant features, and then supplied the remaining relevant attributes to the functional expansion unit of FLANN, which in turn map lower to higher dimensional feature space for constructing a smooth and accurate classifier. In specific, the differential evolution is used to fine tune the weight vector of this network and some trigonometric functions are used in functional expansion unit. The proposed approach is validated with a few benchmarking highly skewed and balanced dataset retrieved from University of California, Irvine (UCI) repository with a range of 5-20% noise. The insightful experimental study signifies...

Classification of Web Logs Using Hybrid Functional Link Artificial Neural Networks

Over the decades, researchers are striving to understand the web usage pattern of a user and are also extremely important for the owners of a website. In this paper, a hybrid analyzer is proposed to find out the browsing patterns of a user. Moreover, the pattern which is revealed from this surge of web access logs must be useful, motivating, and logical. A smooth functional link artificial neural network has been used to classify the web pages based on access time and region. The accuracy and smoothness of the network is taken birth by suitably tuning the parameters of functional link neural network using differential evolution. In specific, the differential evolution is used to fine tune the weight vector of this hybrid network and some trigonometric functions are used in functional expansion unit. The simulation result shows that the proposed learning mechanism is evidently producing better classification accuracy.

Improving Software Reliability Prediction Accuracy Using CRO-Based FLANN

This chapter introduces a novel learning scheme based on chemical reaction optimization (CRO) for training functional link artificial neural network (FLANN) to improve the accuracy of software reliability prediction. The best attributes of FLANN such as capturing the inner association between software failure time and the nearest ‘m’ failure time have been harnessed in this work. Hence, this article combines the best attributes of these two methodologies known as CRO and FLANN to assess the potency in predicting time-to-next failure of software products. The extensive experimental study on a few benchmarking software reliability datasets reveals that the proposed approach fits the historical fault datasets better and more accurately predicts the remaining number of faults than traditional approaches.