LAUSR

Abstract The implementation of Cleaner Production is an important aspect for achieving an industry's sustainable development and is therefore an important instrument for environmental management systems. Due to its special alloy and catalytic properties, the vanadium extraction industry is developing fast, resulting in several environmental problems. The vanadium extraction technology has begun to implement Cleaner Production technologies and practices to tackle environmental problems caused by the production process. However, Cleaner Production evaluation methodologies aimed at small samples have rarely been examined in the literature. This paper proposes data mining methods to the Cleaner Production evaluation system that enable decision-makers to quantitatively evaluate the effectiveness of Cleaner Production. We first completed the indicator framework and criteria of vanadium extraction industry using weak acid leaching. Then established the Cleaner Production assessment model aimed at small samples using Support Vector Machines optimization with Genetic Algorithm. According to the comparison in application, the Genetic Algorithm - Support Vector Machines method is more accurate and efficient than back-propagation artificial neural network. Moreover, performed sensitivity analysis on the input variables and put forward some policy suggestion on the most influencing parameters. As such, this study suggests an effective assessment method of small samples of CP and provides a guideline for enterprise management on the implementation of CP in the industry of vanadium extraction from stone coal.