Muhammad Zubair Khan

Bio: Muhammad Zubair Khan is an academic researcher. The author has contributed to research in topics: Cloud testing & Service quality. The author has an hindex of 1, co-authored 1 publications receiving 14 citations.

Towards Service Evaluation and Ranking Model for Cloud Infrastructure Selection

Abstract: Cloud computing is bringing revolution in the world of information technology by offering on-demand publicly available desired computing resources to the cloud customers. As, there are multiple cloud service providers available in the market and are increasing exponentially, it becomes very difficult for customer to select the most appropriate option from the web repository based on his personalized quality requirements. Also, the dynamic nature of cloud services makes it even difficult for customers to assess resource provider and its service quality ensured in Service Level Agreement. Currently no model exists that allow the customer to evaluate and rank services depending on performance consistency. Under this context, this paper propose Cloud Service Evaluation and Ranking Model (CSERM), which utilizes service measurement index, consistency checking and multiple criteria decision making techniques to evaluate service performance (cloud resource providers) based on user-defined requirements and generate ranked list from where the customer can select the top most ranked option for deploying their applications. Such a model can make significant impact on service performance and compel cloud providers to satisfy their promised quality of services. A case study in this paper describes the whole model implementation.

Cloud service evaluation method-based Multi-Criteria Decision-Making

Abstract: Reviewed 77 papers on cloud service evaluation method-based MCDM.Presented a systematic literature review based on evaluation theory components.Sixteen research deficiencies were identified. A substantial effort has been made to solve the cloud-service evaluation problem. Different Cloud Service Evaluation Methods (CSEMs) have been developed to address the problem. Cloud services are evaluated against multiple criteria, which leads to a Multi-Criteria Decision-Making (MCDM) problem. Yet, studies that assess, analyse, and summarize the unresolved problems and shortcomings of current CSEM-based MCDM are limited. In the existing review studies, only individual parts of CSEMs, rarely the full solution, are reviewed and examined. To investigate CSEMs comprehensively, we present a systematic literature review based on Evaluation Theory, a theory that generalizes six evaluation components, target, criteria, yardstick, data gathering techniques, synthesis techniques, and evaluation process. These six evaluation components and the CSEMs validation approach are the seven dimensions used to assess and analyse 77 papers published from 2006 to 2016. Sixteen research deficiencies were identified. The results confirm that the majority of the studies of the proposed CSEMs were either incomplete or lacked sufficient evidence. This research not only provides the relative strengths and weaknesses of the different CSEMs but also offers a basis for researchers and decision makers to develop improved CSEMs.

Deep Neural Architectures for Medical Image Semantic Segmentation: Review

Abstract: Deep learning has an enormous impact on medical image analysis. Many computer-aided diagnostic systems equipped with deep networks are rapidly reducing human intervention in healthcare. Among several applications, medical image semantic segmentation is one of the core areas of active research to delineate the anatomical structures and other regions of interest. It has a significant contribution to healthcare and provides guided interventions, radiotherapy, and improved radiological diagnostics. The underlying article provides a brief overview of deep convolutional neural architecture, the platforms and applications of deep neural networks, metrics used for empirical evaluation, state-of-the-art semantic segmentation architectures based on a foundational convolution concept, and a review of publicly available medical image datasets highlighting four distinct regions of interest. The article also analyzes the existing work and provides open-ended potential research directions in deep medical image semantic segmentation.

Solving cloud vendor selection problem using intuitionistic fuzzy decision framework

Abstract: This paper presents a new decision-making framework called cloud vendor selector (CVS) for effective selection of cloud vendors by mitigating the challenge of unreasonable criteria weight assignment and improper management of uncertainty. The CVS comprises of two stages where, in the first stage, decision-makers’ intuitionistic fuzzy-valued preferences are aggregated using newly proposed extended simple Atanassov’s intuitionistic weighted geometry operator. Further, in the second stage, criteria weights are estimated by using newly proposed intuitionistic fuzzy statistical variance method and finally, ranking of cloud vendor (CV) is done using newly proposed three-way VIKOR method under intuitionistic fuzzy environment which introduces neutral category along with cost and benefit for better understanding the nature of criteria. An illustrative example of CV selection is demonstrated to show the practicality and usefulness of the proposed framework. Finally, the strength and weakness of the proposal are realized from both theoretic and numeric context by comparison with other methods.

A comparative analysis of prominently used MCDM methods in cloud environment

Abstract: Incloud computing, the selection of an efficient multi-criteria decision-making (MCDM) method (with minimum time complexity and maximum robustness) is a challenging and interesting problem. The time complexity and robustness of a MCDM method depend upon the methodology of evaluating the best alternative (i.e., cloud service). Although numerous MCDM methods are proposed for the quality-of-service based service selection in the cloud, still the issue of selecting the most efficient method remains unresolved. This paper presents a comparative analysis of the prominently used MCDM methods in terms of time complexity and robustness. The MCDM methods are used in the geographical region selection problem for Amazon Web Service cloud, and a comparative analysis of the obtained ranking results is performed. Further, application-specific analysis and sensitivity analysis are performed to ascertain the robustness of ranking methods. Experimental analysis is performed on the large-scale synthetic dataset to get the ranking overhead, i.e., time complexity of different MCDM methods.

ECG Classification using Deep Transfer Learning

Abstract: The state-of-the-art deep neural networks trained on a large amount of data can better diagnose cardiac arrhythmias than cardiologists. However, the requirement of the high-volume training data is not pragmatic. In this research, the identification and classification of three ECG patterns are analyzed from a transfer learning prospect. The features learned from the general image classification are transferred to the time-series signal (ECG) classification using transfer learning. In this research, various modern deep networks trained on the ImageNet database are re-utilized for classifying scalograms (2D representation) of ECG signals. The performance of these deep transfers on the classification of ECG time-series data is then assessed.