STUDY ON ENVIRONMENTAL PROTECTION OF RESOURCE EXHAUSTED CITY BASED ON ECOLOGICAL COMPENSATION PERSPECTIVE

摘要

The influence factors of mine environment are various, and the quantitative evaluation process is susceptible to human intervention. Sparse fuzzy regression model and SVM algorithm can automatically simulate the non-linear relationship among the factors. It is the first time to introduce it into the mine environment assessment. 160 units are selected as training samples, 14 variable indexes of 4 categories, such as physical geography, basic geology, development area and geological environment, are used as input vectors, and unit evaluation scores are used as output vectors. Sparse fuzzy regression model and SVM mine environment assessment model are established respectively. The results show that: Both of the two models can meet the precision requirements of mine environmental assessment. SVM model converges faster than sparse fuzzy regression model, and MSE is smaller than sparse fuzzy regression model, so it is more suitable for mine environment assessment. The quantitative model was applied to the study area, and the evaluation score was divided into four grades, which was consistent with the qualitative evaluation results, and provided a new idea for mine environmental evaluation.