A Machine Learning-Based Approach to Estimate the CPU-Burst Time for Processes in the Computational Grids

The implementation of CPU-Scheduling algorithms such as Shortest-Job-First (SJF) and Shortest Remaining Time First (SRTF) is relying on knowing the length of the CPU-bursts for processes in the ready queue. There are several methods to predict the length of the CPU-bursts, such as exponential averag...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2015 3rd International Conference on Artificial Intelligence, Modelling and Simulation (AIMS) s. 3 - 8
Hlavní autoři: Helmy, Tarek, Al-Azani, Sadam, Bin-Obaidellah, Omar
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.12.2015
Témata:
Artificial intelligence
Computational modeling
CPU Scheduling Algorithm
CPU-Burst
Feature Selection
Machine Learning
Scheduling algorithms
Single machine scheduling
Testing
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The implementation of CPU-Scheduling algorithms such as Shortest-Job-First (SJF) and Shortest Remaining Time First (SRTF) is relying on knowing the length of the CPU-bursts for processes in the ready queue. There are several methods to predict the length of the CPU-bursts, such as exponential averaging method, however these methods may not give an accurate or reliable predicted values. In this paper, we will propose a Machine Learning (ML) based approach to estimate the length of the CPU-bursts for processes. The proposed approach aims to select the most significant attributes of the process using feature selection techniques and then predicts the CPU-burst for the process in the grid. ML techniques such as Support Vector Machine (SVM) and K-Nearest Neighbors (K-NN), Artificial Neural Networks (ANN) and Decision Trees (DT) are used to test and evaluate the proposed approach using a grid workload dataset named "GWA-T-4 Auver Grid". The experimental results show that there is a strength linear relationship between the process attributes and the burst CPU time. Moreover, K-NN performs better in nearly all approaches in terms of CC and RAE. Furthermore, applying attribute selection techniques improves the performance in terms of space, time and estimation.
DOI:10.1109/AIMS.2015.11