Evaluation And Selection Of Green Suppliers In Petrochemical Industry Based On Machine Learning

Based on high-quality development and the realization of the goal of "carbon peaking,carbon neutrality",the petrochemical industry has put forward higher requirements for supplier evaluation and selection of green purchasing.The supplier evaluation index of petrochemical industry lacks unified standard and green index,and the evaluation method is not accurate enough.Therefore,this paper firstly establishes the supplier evaluation index through literature review and whole life cycle theory,then uses the triple bottom line criterion(TBL)and balanced scorecard theory(BSC)to screen the indicators,and obtains the green supplier evaluation index that takes into account economic,social and environmental benefits.In this paper,the combined algorithms of rough set and least squares support vector machine(RS-LSSVM)are used to evaluate the suppliers,and the applicability and superiority of RS-LSSVM combinations are discussed.Rough sets can reduce attributes while ensuring that the sample resolution is unchanged,which can avoid the impact of redundant information on the support vector machine;The supplier is then evaluated using the least squares support vector machine,which has a good generalization ability for small sample and nonlinear problems.The experimental verification is carried out by ROSETTA and MATLAB,and the effectiveness and scientificity of the RSSSVM model are proved by comparative analysis with LSSVM and RS-BP neural network results.The green supplier evaluation system and method model constructed in this paper enrich the existing theoretical results of green supplier evaluation research and promote the practical application of green supplier management in petrochemical industry.