Papers by Harshad Prajapati
Journal of Intelligent Systems, 2017
Clustering is an unsupervised kind of grouping of data points based on the similarity that exists... more Clustering is an unsupervised kind of grouping of data points based on the similarity that exists between them. This paper applied a combination of particle swarm optimization and K-means for data clustering. The proposed approach tries to improve the performance of traditional partition clustering techniques such as K-means by avoiding the initial requirement of number of clusters or centroids for clustering. The proposed approach is evaluated using various primary and real-world datasets. Moreover, this paper also presents a comparison of results produced by the proposed approach and by the K-means based on clustering validity measures such as inter- and intra-cluster distances, quantization error, silhouette index, and Dunn index. The comparison of results shows that as the size of the dataset increases, the proposed approach produces significant improvement over the K-means partition clustering technique.
2011 C Prajapati Analytical Study of parallel and distributed image processing
Journal of Grid Computing, 2015
Due to the high complexity of Grid computing systems, experimentation on a real Grid system is ch... more Due to the high complexity of Grid computing systems, experimentation on a real Grid system is challenging and time consuming. Moreover, deploying a Grid system demands a lot of efforts, money, and skills. Therefore, a simulation based approach of experimentation and research is adopted by many researchers. Many simulation tools are available supporting diverse research studies in the Grid computing area. However, researchers need to choose the most appropriate one for the intended study, and taking that decision requires that the researchers understand all relevant details pertaining to the Grid simulation tools. Therefore, to guide a researcher in choosing a particular Grid simulation tool, we pose important questions that the researcher needs to consider. To answer the posed questions, we provide analysis perspective views of 12 important Grid simulation tools with emphasis on different users. Furthermore, we share our experience of working with SimGrid and GridSim. Our results with 31 comparison criteria on the selected Grid simulation tools would become very
Journal of Computer Networks and Communications, 2014
Bandwidth-aware workflow scheduling is required to improve the performance of a workflow applicat... more Bandwidth-aware workflow scheduling is required to improve the performance of a workflow application in a multisite Grid environment, as the data movement cost between two low-bandwidth sites can adversely affect the makespan of the application. Pegasus WMS, an open-source and freely available WMS, cannot fully utilize its workflow mapping capability due to unavailability of integration of any bandwidth monitoring infrastructure in it. This paper develops the integration of Network Weather Service (NWS) in Pegasus WMS to enable the bandwidth-aware mapping of scientific workflows. Our work demonstrates the applicability of the integration of NWS by making existing Heft site-selector of Pegasus WMS bandwidth aware. Furthermore, this paper proposes and implements a new workflow scheduling algorithm—Level based Highest Input and Processing Weight First. The results of the performed experiments indicate that the bandwidth-aware workflow scheduling algorithms perform better than bandwidth...
2014 Fourth International Conference on Advanced Computing & Communication Technologies
Remote job submission and execution is fundamental requirement of distributed computing done usin... more Remote job submission and execution is fundamental requirement of distributed computing done using Cluster computing. However, Cluster computing limits usage within a single organization. Grid computing environment can allow use of resources for remote job execution that are available in other organizations. This paper discusses concepts of batch-job execution using LRM and using Grid. The paper discusses two ways of preparing test Grid computing environment that we use for experimental testing of concepts. This paper presents experimental testing of remote job submission and execution mechanisms through LRM specific way and Grid computing ways. Moreover, the paper also discusses various problems faced while working with Grid computing environment and discusses their troubleshootings. The understanding and experimental testing presented in this paper would become very useful to researchers who are new to the field of job management in Grid.
Load balancing using process migration for linux based distributed system
2014 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT), 2014
ABSTRACT This paper presents implementation of load balancing mechanism using master-slave model ... more ABSTRACT This paper presents implementation of load balancing mechanism using master-slave model and Berkeley Lab Checkpoint/Restart (BLCR) toolkit. The overall goal is to create a Master-Slave model through which we can migrate processes from highly loaded nodes to some dedicated lightly loaded nodes. The agent running on master node divides total work into equal sub tasks and delegates these sub-tasks to several independent slave nodes. The master node computes final result by aggregating the partial results returned by slaves.
Concise CNN model for face expression recognition
Intelligent Decision Technologies, 2021
Face Expression Recognition (FER) has gained very much attraction to researchers in the field of ... more Face Expression Recognition (FER) has gained very much attraction to researchers in the field of computer vision because of its major usefulness in security, robotics, and HMI (Human-Machine Interaction) systems. We propose a CNN (Convolutional Neural Network) architecture to address FER. To show the effectiveness of the proposed model, we evaluate the performance of the model on JAFFE dataset. We derive a concise CNN architecture to address the issue of expression classification. Objective of various experiments is to achieve convincing performance by reducing computational overhead. The proposed CNN model is very compact as compared to other state-of-the-art models. We could achieve highest accuracy of 97.10% and average accuracy of 90.43% for top 10 best runs without any pre-processing methods applied, which justifies the effectiveness of our model. Furthermore, we have also included visualization of CNN layers to observe the learning of CNN.
ArXiv, 2017
Cloud Computing is a new era of remote computing / Internet based computing where one can access ... more Cloud Computing is a new era of remote computing / Internet based computing where one can access their personal resources easily from any computer through Internet. Cloud delivers computing as a utility as it is available to the cloud consumers on demand. It is a simple pay-per-use consumer-provider service model. It contains large number of shared resources. So Resource Management is always a major issue in cloud computing like any other computing paradigm. Due to the availability of finite resources it is very challenging for cloud providers to provide all the requested resources. From the cloud providers perspective cloud resources must be allocated in a fair and efficient manner. Research Survey is not available from the perspective of resource management as a process in cloud computing. So this research paper provides a detailed sequential view / steps on resource management in cloud computing. Firstly this research paper classifies various resources in cloud computing. It also...
Intelligent Decision Technologies, 2019
In computer vision, recognizing human activity or behavior is a core challenging problem. This ar... more In computer vision, recognizing human activity or behavior is a core challenging problem. This article provides a crisp study of human activity recognition systems in the area of visual surveillance. These systems are used for analysis and understanding of the human behavior. The study starts with the description of various emerging video processing domains, followed by a general process of human action recognition. Then the article covers human detection techniques from images and video. Finally, article also provides a survey of different features and models used in activity recognition systems and an overview of benchmark dataset of video surveillance. From this state-of-the-art survey, researchers can outline promising directions of research.
Intelligent Decision Technologies, 2019
Face detection has been widely studied by researchers. However, detection and extraction of human... more Face detection has been widely studied by researchers. However, detection and extraction of human face features is very important as it plays a vital role in variety of applications involving automated face processing. This article focuses on extraction of face parts such as eyes, nose, lips, mustache, and beard on Indian people, for which we have prepared our own face dataset containing variety in faces, from both urban and rural areas. This study focuses on how a detected face part becomes useful in detecting other face parts. We implement our approaches of detecting face parts and evaluate them on our dataset. We exploit YCbCr color model, Viola Jones technique, landmark detection, and level set evolution technique in our approaches of face part detection and extraction. We found that our approaches are effective on extracting face boundary, eyes, nose, and lips and provide comparable results. Keywords: Image processing, facial features, human face boundary extraction, eye extraction, nose and lip extraction, beard and mustache extraction brows, beard, moustache, chin, etc., and then separat-8 ing these portions for further required processing. Ex-9 traction of human face parts plays an important role in 10 human face analysis [2], visual interpretation, and hu-11 man face recognition [3,4]. Face detection has attracted 12 much interest since a long and has progressed drasti-13 cally over past few decades [5-7]; however, detection 14 of human face parts is of prime importance in a wide 15 variety of applications such as computer vision, facial 16 animation, face recognition, facial expression detec-17 tion, face image database management, etc. A human
Cloud computing has given the new face to the distributed field. Two main issues are discussed in... more Cloud computing has given the new face to the distributed field. Two main issues are discussed in this paper, (I) "the process of finding the efficient virtual machine by using the concept of load balancing algorithm". (II) "Reallocation of the Virtual Machines" i.e. migration of the Virtual Machines when cloud provider is not available with the required Virtual Machines. We have discussed about the different load balancing algorithms which are used for deciding the efficient Virtual Machine for the allocation to the client on demand. While in the second issue is concern we have discuss about different modules available for the migration of Virtual Machines from one source machine to the other target machine. At last discussion about the different simulators available for the cloud are carried out in this paper.
Cutting stock problem: A solution based on novel pattern based chromosome representation using modified GA
2015 International Conference on Circuits, Power and Computing Technologies [ICCPCT-2015], 2015
The cutting stock problem (CSP) is an important problem in class of combinatorial optimization pr... more The cutting stock problem (CSP) is an important problem in class of combinatorial optimization problems because of its NP-hard nature. Cutting the required material from available stock with minimum wastage is a challenging process in many manufacturing industries such as rod industry, paper industry, textile industry, wood industry, plastic and leather manufacturing industry etc. This objective of this research work is to Propose Novel Pattern based chromosome representation using Modified Genetic Algorithm for multiple stock size cutting stock problem (MSSCSP) and Single stock size cutting stock problem (SSSCSP). The main challenge in solving cutting stock problem is to develop chromosome representation for MSSCSP in GA. Moreover, this paper Test results on 20 large dataset with LP, EP and Two-swap algorithm. Our Propose algorithm gives better results than LP in MSSCSP, EP in SSSCSP and Two-swap algorithm in SSSCSP.
Study of Automatic Graph Layout Generation using Genetic Algorithm
Proposed Architecture for Evaluation of Quality of Product Aspect Based on User Reviews
Embedding Bandwidth Monitoring in Ganglia Monitoring System
ABSTRACT The Ability of measuring the Host resource is useful in several domains like Cluster and... more ABSTRACT The Ability of measuring the Host resource is useful in several domains like Cluster and Grid computing. Ganglia monitoring system provides the ability to monitor the resource information of hosts which are present in Grid or Cluster computing environments. Bandwidth is an important aspect in network related applications. Objective of this paper is to embed the bandwidth monitoring functionality in Ganglia monitoring system. This research paper proposes the required architecture for embedding the custom metric module for bandwidth measurement in Ganglia Monitoring System. This architecture measures the bandwidth between any pair of nodes. Pairs are specified into Host- Configuration file and measured bandwidths between all pairs are embedded into Ganglia. Moreover this research paper also presents the results of the embedding bandwidth measurement in Ganglia monitoring system. The presented work would be very useful to researchers who work in the area of resource monitoring in distributed environment.
Grid computing has attracted many researchers over a few years, and as a result many new protocol... more Grid computing has attracted many researchers over a few years, and as a result many new protocols have emerged and also evolved since its inception a decade ago. Grid protocols play major role in implementing services that facilitate coordinated resource sharing across diverse organizations. In this paper, we provide comprehensive coverage of different core Grid protocols that can be used in Global Grid Computing. We establish the classification of core Grid protocols into i) Grid network communication and Grid data transfer protocols, ii) Grid information security protocols, iii) Grid resource information protocols, iv) Grid management protocols, and v) Grid interface protocols, depending upon the kind of activities handled by these protocols. All the classified protocols are also organized into layers of the Hourglass model of Grid architecture to understand dependency among these protocols. We also present the characteristics of each protocol. For better understanding of these protocols, we also discuss applied protocols as examples from either Globus toolkit or other popular Grid middleware projects. We believe that our classification and characterization of Grid protocols will enable better understanding of core Grid protocols and will motivate further research in the area of Global Grid Computing.
International Journal of Computer Applications, 2013
In the last decade, scheduling of Directed Acyclic Graph (DAG) application in the context of Grid... more In the last decade, scheduling of Directed Acyclic Graph (DAG) application in the context of Grid environment has attracted attention of many researchers. However, deployment of Grid environment requires skills, efforts, budget, and time. Although various simulation toolkits or frameworks are available for simulating Grid environment, either they support different possible studies in Grid computing area or takes lot of efforts in molding them to make them suitable for scheduling of DAG application. In this paper, we describe design and implementation of GridSim based ready to use application scheduler for scheduling of DAG application in Grid environment. The proposed application scheduler supports supplying DAG application and configuration of Grid resources through GUI. We also describe implementation of Min-Min static scheduling algorithm for scheduling of DAG application to validate the proposed scheduler. As the proposed DAG application scheduler is based on Java language, it is extensible and portable. Our proposed DAG application scheduling simulator is useful, easy, and time-saver.
2014 Fourth International Conference on Advanced Computing & Communication Technologies
Scheduling in Grid computing has been active area of research since its beginning. However, begin... more Scheduling in Grid computing has been active area of research since its beginning. However, beginners find very difficult to understand related concepts due to a large learning curve of Grid computing. Thus, there is a need of concise understanding of scheduling in Grid computing area. This paper strives to present concise understanding of scheduling and related understanding of Grid computing system. The paper describes overall picture of Grid computing and discusses important subsystems that enable Grid computing possible. Moreover, the paper also discusses concepts of resource scheduling and application scheduling and also presents classification of scheduling algorithms. Furthermore, the paper also presents methodology used for evaluating scheduling algorithms including both real system and simulation based approaches. The presented work on scheduling in Grid containing concise understandings of scheduling system, scheduling algorithm, and scheduling methodology would be very useful to users and researchers.
Analytical study of parallel and distributed image processing
2011 International Conference on Image Information Processing, 2011
ABSTRACT The available literature on parallel and distributed image processing is scattered and n... more ABSTRACT The available literature on parallel and distributed image processing is scattered and not organized for use to beginners. Thus, there is a need of concise understanding of parallel and distributed image processing area. In this paper, we present analysis of parallel and distributed image processing with comprehensive details, so that it becomes very useful to beginners and to those who are new to parallel or distributed image processing field. We present the outcome of our study of parallel and distributed image processing with emphasis on mechanisms, tools/technology/API used, application domains, and ongoing research work. We examine the research issues in parallel and distributed image processing. We also identify some future research directions for distributed image processing. This study provides concise understanding of the parallel and distributed image processing area to the beginners.
2014 IEEE International Advance Computing Conference (IACC), 2014
Since beginning of Grid computing, scheduling of dependent tasks application has attracted attent... more Since beginning of Grid computing, scheduling of dependent tasks application has attracted attention of researchers due to NP-Complete nature of the problem. In Grid environment, scheduling is deciding about assignment of tasks to available resources. Scheduling in Grid is challenging when the tasks have dependencies and resources are heterogeneous. The main objective in scheduling of dependent tasks is minimizing make-span. Due to NP-complete nature of scheduling problem, exact solutions cannot generate schedule efficiently. Therefore, researchers apply heuristic or random search techniques to get optimal or near to optimal solution of such problems. In this paper, we show how Genetic Algorithm can be used to solve dependent task scheduling problem. We describe how initial population can be generated using random assignment and height based approaches. We also present design of crossover and mutation operators to enable scheduling of dependent tasks application without violating dependency constraints. For implementation of GA based scheduling, we explore and analyze SimGrid and GridSim simulation toolkits. From results, we found that SimGrid is suitable, as it has support of SimDag API for DAG applications. We found that GA based approach can generate schedule for dependent tasks application in reasonable time while trying to minimize make-span.
Uploads
Papers by Harshad Prajapati