A Hyper-Heuristic for the Multi-Objective Integration and Test Order Problem

Abstract

Multi-objective evolutionary algorithms (MOEAs) have been efficiently applied to Search-Based Software Engineering (SBSE) problems. However, skilled software engineers waste significant effort designing such algorithms for a particular problem, adapting them, selecting operators and configuring parameters. Hyper-heuristics can help in these tasks by dynamically selecting or creating heuristics. Despite of such advantages, we observe a lack of works regarding this subject in the SBSE field. Considering this fact, this work introduces HITO, a Hyper-heuristic for the Integration and Test Order Problem. It includes a set of well-defined steps and is based on two selection functions (Choice Function and Multi-armed Bandit) to select the best low-level heuristic (combination of mutation and crossover operators) in each mating. To perform the selection, a quality measure is proposed to assess the performance of low-level heuristics throughout the evolutionary process. HITO was implemented using NSGA-II and evaluated to solve the integration and test order problem in seven systems. The introduced hyper-heuristic obtained the best results for all systems, when compared to a traditional algorithm.

A Learning Automata-Based Multiobjective Hyper-Heuristic

“…[30] and [31] incorporated a roulette wheel based heuristic selection mechanism [32] into their multiobjective hyper-heuristic evolutionary algorithm to select low level mutation operators. [33] developed a hyper-heuristic based on two heuristic selection methods (choice function [14] and multi-armed bandit [34]) for choosing from multiple mutation and crossover operators during the search for the multiobjective integration and test order problems [35].…”

Concrete hyperheuristic framework for test case prioritization

“…This approach can improve the selection accuracy but spends much time selecting one for each generation. Guizzo et al also verified two different selection functions for the hyperheuristic algorithm in test order problems. The solution they used in the low level is an algorithm combined with a selection between three crossovers and two mutations.…”

“…There are two evaluation approaches that could be used in the high level according to the related studies: choice function–based approaches and nondominated sorting approaches. The choice function–based approaches evaluate solutions by summarizing the objectives based on a weighting scheme. However, in different testing scenarios, appropriate weights for measure objectives may be difficult to determine.…”

A Stochastic Method for Test Case Selection in Software Testing

“…Results proved that the solution along with its best search algorithm (Random-Weighted Genetic Algorithm (RWGA)) and this will be able to improve current practice by means of bringing down on an average of about 40.6% of the time that had been used for the allocation of resource and the execution of the test cases with an improved usage of test resource by about 37.9% and a fault detection on an average by about 60%. Guizzo et al, [12] had further introduced a new Hyper-heuristic for Integration and Test Order (HITO) issues. This included another new set of some well-designed steps that were based on a total of two different selection functions (the Choice Function and the Multi-Armed Bandit) for the purpose of choosing the ideal heuristic (a combination of both mutation and crossover operations) in each mating.…”