大数据在粮库粮情预测中的应用

Application of big data to forecast of grain situation in grain depot

针对大数据在其他行业产生的聚合增值效应,立足粮情预测问题,分析大数据在我国粮情预测领域缺乏应用的现状。通过建立各级储粮公司数据共享的云服务平台,借助基于部分约束信息的大规模数据集EM概率聚类算法对粮库粮情大数据进行挖掘。围绕粮食储藏安全给出几点指导性意见,为未来粮情预测提供更好的理论技术支撑。指出大数据的应用必将成为提升粮食安全系数的重要手段和途径,同时实现降成本、增效益的目标,与此同时展望未来的发展趋势。

According to the aggregating incremental effect in many industries by big data, the actualities that big data is lacked in the field of grain situation forecast in our country are analyzed. The big data of grain situation in grain depot is collected by establishment of cloud service platform shared among grain storage companies with the help of scalable E M probability clustering algorithm for massive data sets based on partial constraint information. Some opinions about grain storage security are proposed for provi-ding better theory technical support to grain situation forecast in the futurity, and pointing out that the application of big data will become an important means and ways to improve food safety coefficient, so as to realize the goal of cost reduction and benefit increasing, at the same time looking forward to the future development trend.