我国粮食作物保险风险区划的实证研究

Empirical Study of the Insurance Risk Zoning of China's Grain Crops

粮食保险属于政策性农业保险,保险风险区划是全面推进粮食保险的一项先期准备工作。本文认为,产量变化能够测度自然风险对粮食安全的影响,并从这一新的视角建立了粮食安全自然风险影响的评价指标体系,利用系统聚类法、K-均值聚类法和模糊聚类法对我国粮食生产进行了省级保险风险区划划分,并以Fisher判别法、Bayes判别法和逐步判别法进行了回判验证。

Grain insurance belongs to policy-guided agricultural insurance. Insurance risk zoning is an earlier job in preparation for propelling grain crops＇ insurance full-scaly. The paper constructs the evaluation index system of natural risk influence from a new point of view that Output＇s change can measure the influence of natural risks on grain production. And then the insurance risk zoning is done by hierarchical cluster, K-Means cluster and fuzzy cluster analysis approaches and the result is tested by Fisher＇s, Bayes＇ and stepwise analysis methods.