A Boosted Tree Classifier Algorithm based Collaborative Computing Framework for Smart System

Gunasekaran Manogaran, Bharat S. Rawal, Mamoun Alazab

    Research output: Contribution to journalArticlepeer-review


    The smart manufacturing industry relies on technology and automation based services to improve the outcome of reducing human intervention and manual errors. The development of industrial automation solutions using the Internet of Things (IoT) offers a wide range of computing and visualization solutions in a decentralized and pervasive manner. This manuscript introduces a Collaborative Computing Framework (CCF) in a view to improving the performance of the smart industries. Considering the facts of less human intervention and controlled manufacturing processes, CCF relies on harmonized scheduling between different industrial units. The operating schedules of the various functional industrial units are streamlined using this framework, along with the support of boosted tree classifiers. This streamlining provides better computing and task scheduling by exploiting the chained rapport between different functional units. Unambiguously, the production and logistics operation of the smart manufacturing schedules are analyzed using the computing framework to reduce task backlogs. The definite constraint identification and normalization using the classification process helps to minimize latency and re-schedules by 13.41%, 10.89%, and 19.08%, 11.11% respectively for different tasks and their schedules. From the logistics performance assessment, it is seen that the proposed CCF retains less cost factor and delayed instances.

    Original languageEnglish
    Pages (from-to)1082-1090
    Number of pages10
    JournalIEEE Transactions on Network Science and Engineering
    Issue number3
    Early online dateDec 2020
    Publication statusPublished - May 2022


    Dive into the research topics of 'A Boosted Tree Classifier Algorithm based Collaborative Computing Framework for Smart System'. Together they form a unique fingerprint.

    Cite this