University of Kashan Soft Computing Journal 2322-3707 9 2 2021 02 19 A hybrid algorithm based on Gossip architecture using SVM for task scheduling in cloud computing A hybrid algorithm based on Gossip architecture using SVM for task scheduling in cloud computing 84 93 111576 10.22052/scj.2021.242822.0 FA Shiva Razzaghzadeh Department of Computer Engineering, Islamic Azad University, Ardabil Branch, Ardabil, Iran. Parisa Norouzi Kivi Young and Elite Researchers Club, Islamic Azad University, Ardabil Branch, Ardabil, Iran. Babak Panahi Department of Computer Engineering, Islamic Azad University, Ardabil Branch, Ardabil, Iran. Journal Article 2020 10 20 Nowadays, Cloud computing is very important due to its widespread use. The considerable extensiveness and flexibility of cloud computing as well as other advantages led to new challenges such as reliability. Such challenges have more importance among researchers because of existing high number of cloud users. To solve the challenge of reliability, a variety of fault tolerance algorithms were provided in recent years to deal with and fix faults in cloud computing. Despite many efforts, there still exist some problems. This paper aims to present an efficient and new hybrid algorithm where SVM features, and Gossip protocol are used. SVM is used to analyze and categorize virtual machine data based on behavioral patterns. Moreover, Gossip is used to gather data monitoring categories. In the proposed model, processing time, load amount, and reliability are evaluated for better service quality. The simulation results through the CloudSim tools show that proposed method is able to increase the processing speed around 0.65 and decrease makespan to 7.22 seconds. Nowadays, Cloud computing is very important due to its widespread use. The considerable extensiveness and flexibility of cloud computing as well as other advantages led to new challenges such as reliability. Such challenges have more importance among researchers because of existing high number of cloud users. To solve the challenge of reliability, a variety of fault tolerance algorithms were provided in recent years to deal with and fix faults in cloud computing. Despite many efforts, there still exist some problems. This paper aims to present an efficient and new hybrid algorithm where SVM features, and Gossip protocol are used. SVM is used to analyze and categorize virtual machine data based on behavioral patterns. Moreover, Gossip is used to gather data monitoring categories. In the proposed model, processing time, load amount, and reliability are evaluated for better service quality. The simulation results through the CloudSim tools show that proposed method is able to increase the processing speed around 0.65 and decrease makespan to 7.22 seconds. https://scj.kashanu.ac.ir/article_111576_a02d5f54f588d92c90042fd889db6610.pdf