Om Prakash Patel

A review of clustering techniques and developments

This paper presents a comprehensive study on clustering: exiting methods and developments made at various times. Clustering is defined as an unsupervised learning where the objects are grouped on the basis of some similarity inherent among them. There are different methods for clustering the objects such as hierarchical, partitional, grid, density based and model based. The approaches used in these methods are discussed with their respective states of art and applicability. The measures of similarity as well as the evaluation criteria, which are the central components of clustering, are also presented in the paper. The applications of clustering in some fields like image segmentation, object and character recognition and data mining are highlighted.

Quantum Inspired Binary Neural Network Algorithm

In this paper a novel quantum based binary neural network learning algorithm is proposed. It forms three layer network structure. The proposed method make use of quantum concept for updating and finalizing weights of the neurons and it works for two class problem. The use of quantum concept form an optimized network structure. Also performance in terms of number of neurons and classification accuracy is improved. Same is compared with a quantum-based algorithm for optimizing artificial neural networks algorithm (QANN). It is found that there is improvement in the form of number of neurons at hidden layer, number of iterations, training accuracy and generalization accuracy.

Liver Disease Diagnosis Using Quantum-based Binary Neural Network Learning Algorithm

In this paper, a liver disease diagnosis is carried out using quantum-based binary neural network learning algorithm (QBNN-L). The proposed method constructively form the neural network architecture, and weights are decided by quantum computing concept. The use of quantum computing improves performance in terms of number of neurons at hidden layer and classification accuracy and precision. Same is compared with various classification algorithms such as logistic, linear logistic regression, multilayer perceptron, support vector machine (SVM). Results are showing improvement in terms of generalization accuracy and precision.

Quantum Based Neural Network Classifier and Its Application for Firewall to Detect Malicious Web Request

In this paper, a quantum based neural network classifier is designed as Firewall (QNN-F) to detect malicious Web request on the Web. The proposed algorithm forms neural network architecture constructively by adding the hidden layer neuron one by one. The connection weight and threshold of the neuron are decided using the quantum computing concept. Forming a network constructively eliminates the problem of unnecessarily learning of neural network thus save time. The quantum computing concept gives large subspace for selection of appropriate connection weight in evolutionary ways. Also, the threshold value is decided using the quantum computing concept. To increase the performance of system, a Web crawler is also proposed which find objection URLs on the Web according to the objectionable keywords. The proposed algorithm is tested on Web data, to develop a firewall which detects malicious Web request. Extensive testing on 2000 objectionable and non objectionable URLs are done which shows that proposed system works efficiently for detection of objectionable content. To judge the performance of the proposed classifier same dataset has been tested on well-known classifiers Support Vector Machine and Back Propagation neural learning algorithm. The comparison shows that, the QNN-F performs better than other compared algorithms.

A Quantum-Inspired Fuzzy based Evolutionary algorithm for data clustering

In this paper, a Quantum-Inspired Evolutionary Fuzzy C-Means (QIE-FCM) algorithm is proposed. The proposed approach find the true number of clusters and the appropriate value of weighted exponent (m) which is required to be known in advance to perform clustering using Fuzzy C-Means (FCM) algorithm. However, the selection of inappropriate value of m and C may lead the algorithm to converge to the local optima. To address the issue of selecting the appropriate value of m and corresponding value of C. In QIE-FCM, the quantum concept is used in classical computer where m is represented in terms of quantum bits (qubits). The QIE-FCM is based on generations. At each generation (g), quantum gates are used to generate a new value of m. For each generated value of m, FCM algorithm is executed by varying values of C. Then, corresponding to m value appropriate value of C is identified by evaluating local fitness function for generation g. To achieve the global best value of m and C, the global fitness function is evaluated by comparing the local best fitness value in current generation with the best fitness value obtained among all the previous generations. To judge the efficacy of QIE-FCM algorithm, it is compared with two well-known indices and three evolutionary fuzzy based clustering algorithm and their performance is evaluated on four benchmark datasets. Furthermore, the sensitivity of QIE-FCM is also experimentally investigated in this paper.

Advance quantum based binary neural network learning algorithm

In this paper a quantum based binary neural network algorithm is proposed, named as Advance Quantum based Binary Neural Network Learning Algorithm (AQ-BNN). It forms neural network structure constructively by adding neurons at hidden layer. The connection weights and separability parameter are decided using quantum computing concept. Constructive way of deciding network not only eliminates over-fitting and underfitting problem but also saves time. The connection weights have been decided by quantum way, it gives large space to select optimal weights. A new parameter that is quantum separability is introduced here which find optimal separability plane to classify input sample in quantum way. For each connection weights it searches for optimal separability plane. Thus the best separability plane is found out with respect to connection weights. This algorithm is tested with three benchmark data set and produces improved results than existing quantum inspired and other classification approaches.

On Construction of Multi-class Binary Neural Network Using Fuzzy Inter-cluster Overlap for Face Recognition

In this paper, we propose a Novel Fuzzy-based Constructive Binary Neural Network (NF-CBNN) learning algorithm for multi-class classification. Our method draws a basic idea from Expand and Truncate Learning (ETL), which is a neural network learning algorithm. The proposed method works on the basis of unique core selection, and it guarantees to improve the classification performance by handling overlapping issues among data of various classes by using inter-cluster overlap. To demonstrate the efficacy of NF-CBNN, we tested it on the ORL face data set. The experimental results show that generalization accuracy achieved by NF-CBNN is much higher as compared to the BLTA classifier.

Quantum-inspired evolutionary approach for selection of optimal parameters of fuzzy clustering

Recently, Fuzzy c-Means (FCM) algorithm is most widely used because of its efficiency and simplicity. However, FCM is sensitive to the initialization of fuzziness factor (m) and the number of clusters (c) due to which it easily trapped in local optima. A selection of these parameters is a critical issue because an adverse selection can blur the clusters in the data. In the available fuzzy clustering literature, cluster validity index is used to determine the optimal number of clusters for the dataset, but these indexes may trap into the local optima due to the random selection of m. From the perspective of handling local optima problem, we proposed a hybrid fuzzy clustering approach referred as quantum-inspired evolutionary fuzzy c-means algorithm. In the proposed approach, we integrate the concept of quantum computing with FCM to evolve the parameter m in several generations. The evolution of fuzziness factor (m) with the quantum concept aims to provide the better characteristic of population diversity and large search space to find the global optimal value of m and its corresponding value of c. Experiments using three real-world datasets are reported and discussed. The results of the proposed approach are compared to those obtained from validity indexes like

Quantum Based Learning with Binary Neural Network

V_{CWB}