Research On The Eco-efficiency Evaluation And Forecasting Model Based On KPCA

With the development of social economy,the contradiction between economic growth and resource environment is becoming more and more obvious,which has become a great obstacle on the road of sustainable development in our country.In order to coordinate the relationship between economy and resource environment,the research on ecological efficiency is of great significance.Firstly,according to the selection of kernel principal component analysis method of kernel function and kernel parameters,we used nonlinear regression and information entropy,UCI empirical formula and statistical yearbook original data set information entropy and fitting parameters based on kernel function22?(10)(28)291.11274-410.21300612.9958-xx.Secondly,we summaried the relevant literature,and summarized in the comprehensive objectives and principles of index selection based on the existing research at home and abroad and the research on eco-efficiency,combined with the current economic development and resources and the environment.The original ecological efficiency was improved to forecast the evaluation index system.We used kernel principal component analysis method,data envelopment analysis method and the classical ecological efficiency evaluation model,the KPCA-DEA composite evaluation model was established.Then,we used kernel principal component analysis method and BP neural network prediction method to improve the classical prediction model,the KPCA-BP composite prediction model.Finally,the evaluation model of KPCA-DEA composite and KPCA-BP composite prediction model is analyzed.The results showed that the comprehensive technical efficiency evaluation model for DEA value of 14 to 1,the overall technical efficiency of PCA-DEA model was 1 with 12,while the comprehensive technical efficiency model of KPCA-DEA value 6 was 1.The results showed that from the comprehensive technical efficiency evaluation model,composite KPCA-DEA is superior to PCADEA and DEA.The evaluation of ecological efficiency has certain advantages and feasibility.An empirical analysis of the composite prediction model of KPCA-BP showed that the relative error of KPCA-BP was 1.56% PCA-BP,the relative error was 17.15%,the relative error of BP model was20.75%.The results showed that the KPCA-BP composite prediction model has certain advantages in prediction.This study enriched the evaluation and forecasting methods of ecological efficiency,and provided a reference for the evaluation and prediction of ecological efficiency.