Performance Evaluation Research of Ecological Civilization Policy Based on Data Envelopment Analysis and Artificial Neural Networks Model

Abstract

Ecological civilization is a new form of human civilization that solves the contradiction between development and environment. Because the ecological civilization policy system has is complex and nonlinear, this paper combines Artificial Neural Network technology with the Data Envelopment Analysis(DEA) method and proposes the hybrid DEA-Artificial Neural Network (DEANN) model to estimate the impact of ecological civilization policy on economic development. The model uses regional eco-efficiency evaluated by DEA as one of the inputs for the neural network; the neural network integrates the total input and output results to provide a quantitative estimate of the ecological civilization policy in different provincial regions from 2003-2016. By training and examining the evaluation results with the DEANN model, this paper tries to predict the impact of 13th Five-Year Total Emission Reduction Policy on economic development. The empirical research results and suggestions can improve the ecological civilization policy performance by negative feedback mechanisms.