Quantitative Assessment and Prediction for Industrial Economy Development Stress on Eco-Environment

The industrial economy has a tremendous impact on the ecological environment. A three-layer artificial neural network (ANN) models were developed to predict industrial economic development stress on eco-environment. The results showed that ESI was the highest in 1997, whereas ESI was the lowest in 1998. ESI in different years has fluctuated, but all belong to fourth level, representing higher eco-environment pressure. Standardization transformation is the best among square root transformation, standardization transformation and natural logarithm transformation. From a short-term forecasting viewpoint, the ANN based on BP algorithm yielded the lowest of average MAPE from 1997 to 2004. The artificial neural network models are useful tool for evaluation and prediction for industrial economy development stress on eco-environment and may be applicable to prediction of other ecological regions.