Bhabani Shankar Prasad Mishra

Parallel Computing Environments: A Review

Abstract This article presents a review of various hardware and software parallel computing environments so far developed for the implementation of parallel algorithms to solve complex and computationally dominated optimization problems. These parallel environments are composed of parallel programming tools, performance evaluation tools, debuggers and optimization libraries. The environments can be classified based on the ease of implementation, hardware independence, ease to understand, and guarantee of performance.

Gray-level image enhancement using differential evolution optimization algorithm

Differential Evolution (DE) algorithm represent an adaptive search process for solving engineering and machine learning optimization problems. This paper presents an attempt to demonstrate its adaptability and effectiveness for searching global optimal solutions to enhance the contrast and detail in a gray scale image. In this paper contrast enhancement of an image is performed by gray level modification using parameterized intensity transformation function that is considered as an objective function. The task of DE is to adapt the parameters of the transformation function by maximizing the objective fitness criterion. Experimental results are compared with other enhancement techniques, viz. histogram equalization, contrast stretching and particle swarm optimization (PSO) based image enhancement techniques.

Parallel multi-objective genetic algorithms for associative classification rule mining

Association and classification rule mining are two well-known techniques used in data mining. The integrated approach is known as associative classification rule mining (ACRM), which has helped in developing a compact and efficient classifier for the classification of unknown samples. In this paper, we treated the ACRM as a multi-objective problem and applied the Parallel Multi-objective Genetic Algorithms (PMOGAs) to solve it. ACRM is associated with two phases like rule extraction and rule selection. As ACRM is a multi-objective problem so by applying PMOGA on it we can optimize the measures like support and confidence of association rule mining to extract classification rules in rule extraction phase and in rule selection phase a small number of rules are targeted from the extracted rules to design an accurate and compact classifier, which can maximize the accuracy of the rule set and minimize their complexity. Experiments are conducted on UCI data set by using MOGA and PMOGA. Finally the computational results are analyzed and concluded that the PMOGA for multi-objective rule selection generates a Pareto optimal rule sets with a compact set of classification rules in less time vis-a-vis to MOGA without severely degrading their classification accuracy.

Parallel Single and Multiple Objectives Genetic Algorithms: A Survey

This paper critically reviews the reported research on parallel single and multi-objective genetic algorithms. Many early efforts on single and multi-objective genetic algorithms were introduced to reduce the processing time needed to reach an acceptable solution. However, some parallel single and multi-objective genetic algorithms converged to better solutions as compared to comparable sequential single and multiple objective genetic algorithms. The authors review several representative models for parallelizing single and multi-objective genetic algorithms. Further, some of the issues that have not yet been studied systematically are identified in the context of parallel single and parallel multi-objective genetic algorithms. Finally, some of the potential applications of parallel multi-objective GAs are discussed.

An Automated Ear Localization Technique Based on Modified Hausdorff Distance

Localization of ear in the side face images is a fundamental step in the development of ear recognition based biometric systems. In this paper, a well-known distance measure termed as modified Hausdorff distance (MHD) is proposed for automatic ear localization. We introduced the MHD to decrease the effect of outliers and allowing it more suitable for detection of ear in the side face images. The MHD uses coordinate pairs of edge pixels derived from ear template and skin regions of the side face image to locate the ear portion. To detect ears of various shapes, ear template is created by considering different structure of ears and resized it automatically for the probe image to find exact location of ear. The CVL and UND-E database have side face images with different poses, inconsistent background and poor illumination utilized to analyse the effectiveness of the proposed algorithm. Experimental results reveal the strength of the proposed technique is invariant to various poses, shape, occlusion, and noise.

Swarm Intelligence in Multiple and Many Objectives Optimization: A Survey and Topical Study on EEG Signal Analysis

This paper systematically presents the Swarm Intelligence (SI) methods for optimization of multiple and many objective problems. The fundamental difference of Multiple and Many Objective Optimization problems have been studied very rigorously. The three forefront swarm intelligence methods, i.e., Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO), and Artificial Bee Colony Optimization (ABC) has been deeply studied to understand their ways of solving multiple and many objective problems distinctly. A pragmatic topical study on the behavior of real ants, bird flocks, and honey bees in solving EEG signal analysis completes the survey followed by discussion and extensive number of relevant references.

Detection and Reduction of Impulse Noise in RGB Color Image Using Fuzzy Technique

A concept of impulse noise reduction method for an RGB color image with a fuzzy detection phase is introduced and a fuzzy de-noising procedure is used to filter the color image. In this paper, each color component is correlated to the other two corresponding color components to overcome the color disorder on edge and texture pixel. Here the filtering technique is only applied to noisy pixel, detected by fuzzy technique, while preserving the color and edge sharpness. Experimental results show that the proposed method provides noteworthy improvement on other non-fuzzy and fuzzy filters.

Weapon target assignment problem: multi-objective formulation, optimisation using MOPSO and TOPSIS

In this paper, a static weapon target assignment problem is studied by optimising the conflicting criteria like shooting failure and number of weapons used to destroy the targets. The inherent intractability and conflicting objectives of this problem motivated us to use multi-objective particle swarm optimisation (MOPSO) to uncover the true Pareto front. We first employ the MOPSO to uncover the Pareto front. Secondly, a ranking method called techniques for order preference by similarity to ideal solution (TOPSIS) is used to sort the non-dominated solutions by the preference of decision maker (DM). A numerical experiment on two test cases has been conducted to realise the efficacy of the method. The experimental work is offering large number of solutions in the Pareto front, which may create problem to DM for effective decision. Therefore, by TOPSIS a prioritised set of non-dominated solutions is provided to DM, which fits the preference under different situations.

Fog Assisted Cloud Computing in Era of Big Data and Internet-of-Things: Systems, Architectures, and Applications

This book chapter discusses the concept of edge-assisted cloud computing and its relation to the emerging domain of “Fog-of-things (FoT)”. Such systems employ low-power embedded computers to provide local computation close to clients or cloud. The discussed architectures cover applications in medical, healthcare, wellness and fitness monitoring, geo-information processing, mineral resource management, etc. Cloud computing can get assistance by transferring some of the processing and decision making to the edge either close to client layer or cloud backend. Fog of Things refers to an amalgamation of multiple fog nodes that could communicate with each other with the Internet of Things. The clouds act as the final destination for heavy-weight processing, long-term storage and analysis. We propose application-specific architectures GeoFog and Fog2Fog that are flexible and user-orientated. The fog devices act as intermediate intelligent nodes in such systems where these could decide if further processing is required or not. The preliminary data analysis, signal filtering, data cleaning, feature extraction could be implemented on edge computer leading to a reduction of computational load in the cloud. In several practical cases, such as tele healthcare of patients with Parkinson’s disease, edge computing may decide not to proceed for data transmission to cloud (Barik et al., in 5th IEEE Global Conference on Signal and Information Processing 2017, IEEE, 2017) [4]. Towards the end of this research paper, we cover the idea of translating machine learning such as clustering, decoding deep neural network models etc. on fog devices that could lead to scalable inferences. Fog2Fog communication is discussed with respect to analytical models for power savings. The book chapter concludes by interesting case studies on real world situations and practical data. Future pointers to research directions, challenges and strategies to manage these are discussed as well. We summarize case studies employing proposed architectures in various application areas. The use of edge devices for processing offloads the cloud leading to an enhanced efficiency and performance.

Workflow Scheduling in Cloud Computing Environment Using Bat Algorithm

The data handling and processing capabilities of current computing systems are increasing, owing to applications involving the bigger size of data. Hence, the services have become more expensive. To maintain the popularity of cloud environment due to less cost for such requirements, an appropriate scheduling technique is essential, which will decide what task will be executed on which resource in a manner that will optimize the overall costs. This paper presents an application of the Bat Algorithm (BA) for scheduling a workflow application (i.e., a data intensive application), in cloud computing environment. The algorithm is successfully implemented and the results compared with two popular existing algorithms, namely Particle Swarm Optimization (PSO) and Cat Swarm Optimization (CSO). The proposed BA algorithm gives an optimal processing cost with better convergence and fair load distribution.