The following information was submitted:
Transactions: WSEAS TRANSACTIONS ON SYSTEMS AND CONTROL
Transactions ID Number: 32-201
Full Name: Daniel Vladušič
Position: Doctor (Researcher)
Age: ON
Sex: Male
Address: Pot za Brdom 100
Country: SLOVENIA
Tel:
Tel prefix:
Fax:
E-mail address: daniel.vladusic@xlab.si
Other E-mails: daniel.vladusic@gmail.com
Title of the Paper: Using the black-box approach with machine learning methods in order to improve job scheduling in GRID environments
Authors as they appear in the Paper: Daniel Vladušič
Email addresses of all the authors: daniel.vladusic@xlab.si
Number of paper pages: 12
Abstract: This article focuses on mapping jobs to resources with use of off-the-shelf machine learning methods. The machine learning methods are used in the black-box manner, having a wide variety of parameters for internal cross validation. In the article we focus on two sets of experiments, both constructed with a goal of congesting the system. In the first part, the machine learning methods are used as assistants for the base resource selection algorithms, while in the second part, they are used to directly select the resource for job execution. We set up two experiments with different objectives. In the first experiment, the objective was to maximize the number of jobs finished within the deadline, while in the second experiment, the objective was to minimize the job tardiness. Finally, in the first part we were changing resource selections of simple job dispatching algorithms, while in the second part, we were employing the optimization scheme of the ALEA algorithm. Th!
e results show that even with a such black-box approach of the off-the-shelf machine learning methods, we can achieve improvements or at least get comparable results.
Keywords: Scheduling, Dispatching jobs in GRIDs, GRID, machine learning
EXTENSION of the file: .pdf
Special (Invited) Session:
Organizer of the Session:
How Did you learn about congress:
IP ADDRESS: 193.77.81.82
