摘要
随着农业信息化的发展,农业生产领域积累了大量数据,这些数据由于受到农业生产成本和生产环境的影响包含有一些可用性不高的规模数据,如何提高海量农业数据的可用性以及发现劣质数据的潜在价值是我们需要研究的问题。文章通过对常用机器学习框架和平行学习框架在处理大数据的优劣进行比较,认为平行学习框架在处理大规模数据和数据预测方面有着显著的优势。其次在平行学习框架的基础上提出一种农业大数据预测系统来提高农业大数据的可用性和挖掘劣质数据的潜在价值。最后分析了大数据时代背景下农情发展的方向和趋势。
With the development of agricultural informatization, a large number of data have been accumulated in the field of agricultural production. Due to the impact of agricultural production costs and production environment, these massive data contain some scale data with low availability. How to improve the availability of massive agricultural data and to find the potential value of inferior data are the problems we need to study. This paper compares the advantages and disadvantages of common machine learning frameworks and parallel learning frameworks in dealing with big data. And next, it is concluded that parallel learning framework has significant advantages in dealing with large scale data and data prediction. Secondly, based on parallel learning framework, an agricultural big data prediction system is proposed to improve the availability and availability of large agricultural data. Finally, the direction and trend of agricultural development under the background of big data age are analyzed.
作者
吴文平
潘正高
卢彪
Wu Wenping;Pan Zhenggao;Lu Biao(Suzhou University, Suzhou, Anhui 23400)
出处
《绥化学院学报》
2018年第5期158-160,共3页
Journal of Suihua University
基金
宿州学院横向合作项目(编号:2017hx0001)
教育部2017年第二批产学合作协同育人项目(编号:20170203009)
关键词
平行学习
农业
大数据
异常预测
可用性
parallel learning
agriculture
big data
anomaly prediction
availability