Using combination recurrent neural network and fuzzy time series for data envelopment analysis (DEA)

Abstract

Data envelopment analysis (DEA) is a mathematical programming based method to measure empirically the efficiency and productivity of operating units using multiple inputs to secure multiple outputs. Typically the inputs and the output are incommensurate. In large data set, discussion regarding the forecast and output calculating of decision making units to measure their efficiency is important task specially. In this paper, one new hybrid method of two old forecasting models (fuzzy time series and recurrent neural network), that about data envelopment analysis has been considered, is used in order to get more accurate results than using each of methods individually. In the end of paper, each of methods (fuzzy time series, recurrent neural network, and hybrid method) on large data set of decision making units has been used and the results have been compared to each other.