000			04501ntm a2200373 i 4500
999			_c94984 _d94990
003			MY-KuUP
005			20251125105821.0
006			t||||fr|||| 000 0
007			ta
008			201123t2 2 m a|||fram|| 000 0 eng d
020			_aTHE0008993(Local) _qhardback
040			_aUMP _beng _cUMP _erda
090			_aFKOM .H37 2020 r Thesis
100	0		_aHasneeza Liza Zakaria, _eauthor.
245	1	0	_aElitist hybrid migrating birds optimization and genetic algorithm based strategy for T-way test suite generation / _cHasneeza Liza Zakaria
264		1	_aKuantan, Pahang : _bUMP, _c2020
264		4	_c© 2020
300			_axiv, 132 pages : _billustrations (some color) ; _c30 cm. + _e1 CD-ROM
336			_atext _2rdacontent
336			_atext _2rdacontent
337			_aunmediated _2rdamedia
337			_acomputer _2rdamedia
338			_avolume _2rdacarrier
338			_acomputer disc _2rdacarrier
347			_atext file _bPDF _2rda
500			_aFaculty of Computing
502			_aThesis (Doctor of Philosophy) -- Universiti Malaysia Pahang – 2020
504			_aIncludes bibliographical references
520	3		_aSoftware is essential in our multifaceted lifestyle today, from everyday usage to space exploration. Testing is a crucial part of software development as it determines whether the developed software is met its requirements. The ever increasing line of codes makes it impossible to test the software exhaustively. Traditional testing methods such as equivalence partitioning, boundary value analysis and decision tables are well known methods to reduce test size. Equivalence partitioning assumes that all data in a class are equally partitioned. Furthermore, equivalence partitioning must be complemented with boundary value analysis to ensure enough testing at all the input boundaries. Decision table incorporates testing of the flow of the program. While all these traditional testing methods are useful, they do not deal with interaction testing of inputs. To deal with interaction testing. the adoption of t-way testing, where t indicates the interaction strength, is known to be effective as far as sampling of the tests in a systematic manner. Derived from mathematical object called covering arrays, many t-way strategies have been developed utilizing different approaches such as algebraic, general computational as well as meta-heuristics. Recently, the adoption of meta-heuristics as the backbone of t-way strategies is becoming popular owing to its effectiveness in terms of generating the most minimal test suite sizes. Although useful, much existing meta-heuristic based strategies have not sufficiently explored the adoption of more than one meta-heuristic to perform the search (termed hybridization). Specifically, the exploration and exploitation of existing strategies has been limited based on the (local and global) search operators derived from a single meta-heuristic algorithm. In this case, choosing a proper combination of search operators can be the key for achieving good performance (as hybridization can capitalize on the strengths and address the deficiencies of each individual algorithm in a collective and synergistic manner). Addressing the aforementioned issues, this research proposes the development and implementation of hybrid t-way strategy based Migrating Birds Optimization Algorithm (MBO) and Genetic Algorithm (GA) with elitism, termed Elitist Hybrid MBO-GA. This is to solve the MBO’s early convergence problem with GA’s ability to diversify solutions. The Elitist Hybrid MBO-GA is then compared with the original MBO strategy and several other benchmarked strategies. The proposed strategy serves as our research conduit to investigate the effectiveness of hybrid meta-heuristics for t-way test generation. The Elitist Hybrid MBO-GA manages to get the similar best result with other benchmarked strategies in 17 experiments. The Elitist Hybrid MBO-GA also outperforms other strategies in 8 experiments. Thus, the Elitist Hybrid MBO-GA gets a good result for 25 out of 33 experiments that is 75% of the experiments. Furthermore, the statistical analysis shows 87.5% statistical significance based on the pair comparison of Wilcoxon signedrank. Therefore, this study concludes that that Elitist Hybrid MBO-GA is a useful strategy for generating t-way test suite generation.
610	2	0	_aFaculty of Computing _xDissertations
650		0	_aUniversities and colleges _xDissertations
650		0	_aTheses
942			_2lcc _cTHESIS