基于支持向量机的农机装备水平差异分类研究

The level classification of agricultural equipment based on support vector machine

针对黑龙江垦区各农场农机装备水平不平衡差异问题,采用支持向量机多类分类和主成分分析相结合的方法进行研究。将农机装备水平分为3个等级水平,从总量、速度和均量3方面选取10个评价指标;采用主成分分析法确定5个主要合成指标;建立多类分类支持向量机新模型,把新模型转化成一个互补问题,利用Lagrangian隐函数进一步转化成一个强凸的无约束优化问题,采取快速牛顿算法进行求解;利用实证调研数据,从发展差异度的角度对黑龙江垦区98个农场农机装备水平的差异进行多指标分析,在分类的准确度和训练速度方面都有很好的表现。

In view of the unbalanced development of agricultural equipment in the reclamation farms in Heilongjiang,we use the methods of multi-class support vector machines and principal component analysis to evaluate the differences of agricultural machinery equipment level in these farms.The level of agricultural equipment is divided into three grades.We select ten evaluation indexes from the aspect of the total equipment amount,development speed and volume index.Five major synthetic indicators are identified by using principal component analysis with a new formulation based on the K-SVCR method.We then transform it as a complementarity problem and further a strongly convex unconstrained optimization problem by using the implicit Lagrangian function.Then a fast Newton algorithm with global and finite termination properties is established for solving the resulting optimization problem.For measuring the developmental condition of the agricultural equipment in 98 farms,evaluating and analyzing the comprehensive differences of agricultural equipment level o in each farm,we put forward the policy recommendations to coordinate the development of farm machinery and equipment in each farm in Heilongjiang Reclamation area.