HYBRID GENETIC ALGORITHMS AND ARTIFICIAL NEURAL NETWORKS APPROACH TO NONLINEAR COST ALLOCATION FOR ACTIVITY-BASED COSTING

摘要

Activity-based costing (ABC) has received extensive attention during the past decade.rnHowever, it has some limitations. The first limitation is that ABC does not have general criteria to selectrnrelevant cost drivers. Second, ABC assumes linearity between the uses of activities and the assignedrnquantities of indirect cost. When cost behavior is nonlinear, ABC may distort product costs. This paperrnproposes hybrid artificial intelligence techniques to resolve above two limitations. This study usesrngenetic algorithms (GA) to identify optimal or near-optimal cost drivers. In addition, this paper employsrnartificial neural networks (ANN) to allocate indirect costs with nonlinear behavior. Empirical resultsrnshow that the proposed approach outperforms conventional model.