Litcius/Paper detail

Model Specification Searches in Structural Equation Modeling with a Hybrid Ant Colony Optimization Algorithm

Zeyuan Jing, Huan Kuang, Walter L. Leite, Katerina M. Marcoulides, Charles L. Fisk

2022Structural Equation Modeling A Multidisciplinary Journal16 citationsDOI

Abstract

Model specification is a crucial aspect of structural equation modeling (SEM), since a misspecified model may lead to biased parameter estimation and result in inaccurate conclusions. We propose the Hybrid Ant Colony Optimization Algorithm (hACO), an improved metaheuristic algorithm to conduct model specification searches in SEM. This data mining algorithm combines aspects of the Ant Colony Optimization algorithm with the Tabu search algorithm to increase both accuracy and efficiency. A Monte Carlo simulation study showed that the hACO algorithm provided accurate and efficient SEM specification searches across all designed simulation conditions. The hACO algorithm can help applied researchers conduct specification searches while avoiding potential model misspecifications.

Topics & Concepts

Ant colony optimization algorithmsTabu searchMetaheuristicStructural equation modelingComputer scienceAlgorithmParallel metaheuristicSpecificationMonte Carlo methodHybrid algorithm (constraint satisfaction)Ant colonyMathematical optimizationMeta-optimizationArtificial intelligenceMachine learningMathematicsStatisticsConstraint satisfactionProbabilistic logicConstraint logic programmingAdvanced Text Analysis TechniquesMulti-Criteria Decision MakingAdvanced Multi-Objective Optimization Algorithms
Model Specification Searches in Structural Equation Modeling with a Hybrid Ant Colony Optimization Algorithm | Litcius