A hybrid kidney algorithm strategy for combinatorial interaction testing problem
Combinatorial Interaction Testing (CIT) generates a sampled test case set (Final Test Suite (FTS)) instead of all possible test cases. Generating the FTS with the optimum size is a computational optimization problem (COP) as well as a Non-deterministic Polynomial hard (NP-hard) problem. Recent studi...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/38470/1/A%20hybrid%20kidney%20algorithm%20strategy%20for%20combinatorial%20interaction%20testing%20problem.ir.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Combinatorial Interaction Testing (CIT) generates a sampled test case set (Final Test Suite (FTS)) instead of all possible test cases. Generating the FTS with the optimum size is a computational optimization problem (COP) as well as a Non-deterministic Polynomial hard (NP-hard) problem. Recent studies have implemented hybrid metaheuristic algorithms as the basis for CIT strategy. However, the existing hybrid metaheuristic-based CIT strategies generate a competitive FTS size, there is no single CIT strategy can overcome others existing in all cases. In addition, the hybrid metaheuristic-based CIT strategies require more execution time than their own original algorithm-based strategies. Kidney Algorithm (KA) is a recent metaheuristic algorithm and has high efficiency and performance in solving different optimization problems against most of the state-of-the-art of metaheuristic algorithms. However, KA has limitations in the exploitation and exploration processes as well as the balancing control process is needed to be improved. These shortages cause KA to fail easily into the local optimum. This study proposes a low-level hybridization of KA with the mutation operator and improve the filtration process in KA to form a recently Hybrid Kidney Algorithm (HKA). HKA addresses the limitations in KA by improving the algorithm's exploration and exploitation processes by hybridizing KA with mutation operator, and improve the balancing control process by enhancing the filtration process in KA. HKA improves the efficiency in terms of generating an optimum FTS size and enhances the performance in terms of the execution time. HKA has been adopted into the CIT strategy as HKA based CIT Strategy (HKAS) to generate the most optimum FTS size. The results of HKAS shows that HKAS can generate the optimum FTS size in more than 67% of the benchmarking experiments as well as contributes by 34 new optimum size of FTS. HKAS also has better efficiency and performance than KAS. HKAS is the first hybrid metaheuristic-based CIT strategy that generates an optimum FTS size with less execution time than the original algorithm-based CIT strategy. Apart from supporting different CIT features: uniform/VS CIT, IOR CIT as well as the interaction strength up to 6, this study also introduces another recently variant of KA which are Improved KA (IKA) and Mutation KA (MKA) as well as new CIT strategies which are IKA-based (IKAS) and MKA-based (MKAS). |
|---|