GENETIC ALGORITHM, DATA ENVELOPMENT ANALYSIS AND GLOBAL PERFORMANCE

摘要

A relevant question for the strategic planning is how to identify an adequate set of indicators and obtain the global performance of the firm relative to the competence and the main objectives of the strategy. In this paper, we evaluate the use of a genetic algorithm (GA) to resolve the problem for a larger number of firms and indicators, correcting a deficiency of the Data Envelopment Analysis (DEA). We proceed a empirical comparison between the GA and a DEA technique in such a way to validate it. The algorithm proposed showed robustness and consistency in their results and no restrictions on the computational complexity of the addressed problem. This allows its use in calculating the technical efficiency for any real scenario of possible firms.