|Mueller Warrant, George|
Submitted to: European Journal of Operations Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/24/2007
Publication Date: 11/7/2007
Citation: Whittaker, G.W., Confesor, R., Griffith, S.M., Fare, R., Grosskopf, S., Steiner, J.J., Mueller Warrant, G.W., Banowetz, G.M. 2007. A hybrid genetic algorithm for multiobjective problems with activity analysis-based local search. European Journal of Operations Research.193:195-203. Interpretive Summary: For the assessment of the effects of conservation practices, an economic model was required that was integrated with a biophysical model of agricultural production and the environment. The effects of conservation practices are varied, including producers profits and environmental effects, and some policies are more cost-effective than others. Therefore, or model had to take account of multiple objectives, that sometimes conflicted. The model that we developed integrates an economic model specified with data envelopment analysis with a biophysical model using a genetic algorithm to find the best solutions. The method finds the best solutions for several objectives at once, giving stakeholders a complete set of information about the trade-offs among objectives. We describe a parallel computing approach to computation of the genetic algorithm, and apply the algorithm to evaluation of an input tax to regulate pollution from agricultural production.
Technical Abstract: The objective of this research was the development of a method that integrated a data envelopment analysis economic model of production with a biophysical model, with optimization over multiple objectives. We specified a hybrid genetic algorithm using DEA as a local search method, and NSGA-II for calculation of the multiple objective Pareto optimal set. We describe a parallel computing approach to computation of the genetic algorithm, and apply the algorithm to evaluation of an input tax to regulate pollution from agricultural production.