Degree

Master of Science in Computer Science

Faculty / School

Faculty of Computer Sciences (FCS)

Department

Department of Computer Science

Date of Submission

2016-01-01

Advisor

Dr. Sajjad Haider

Project Type

MSCS Survey Report

Abstract

Nature Inspired Meta-heuristics (NIM) algorithms have made a profound contribution to the field of optimization since they were initiated in the late 190 m se popularity in the last decade. The aim of this research survey is to study and compare the performance of different types of NIM algorithms by the means of evaluating them over several benchmark functions and compare the "best so far" value obtained by those algorithms. Every algorithm was run multiple times to ensure diversity in results and the measure for comparison for amongst the algorithms chose are: (1) the mean of test so far" value obtained for all re-runs of the algorithm, (2) the standard deviation of the "best so far" value obtained for all re-runs of the algorithm, (3) the maximum value obtained by the algorithm amongst all the re-runs and (4) the average runtime of all re-runs of the algorithm to obtain the "best so far" value. As all the algorithms have been implemented for this survey, these algorithms have been combined in a form of a toolkit so that the user can see the performance of the algorithms while changing the parameters themselves.

The full text of this document is only accessible to authorized users.

Share

COinS