摘要
面对当前海洋环境污染局势,有效地对海洋环境进行预测具有重大的现实意义。针对当前的众多海洋数据预测模型采用单步长时间序列预测问题,本文采用平滑支持向量机,对海洋环境数据进行多步长时间序列预测,其预测时间尺度为8小时。实验结果表明,该模型具有良好的回归性能,可以有效地多时间尺度预测海洋环境数据。
Facing the current pollution situation of the marine environment,it is of great practical significance to effectively predict the marine environment.Aiming at many current ocean data prediction models that use single-step time series prediction prob⁃lems,this paper uses a smooth support vector machine to perform multi-step time series predictions on marine environmental data.The prediction time scale is 8 hours.Experimental results show that the model has good regression performance and can effectively predict marine environment data on multiple time scales.
作者
李志刚
王宁
LI Zhi-Gang;WANG Ning(North China University of Science and Technology,Tangshan 063200,China)
出处
《电脑知识与技术》
2020年第16期28-30,共3页
Computer Knowledge and Technology
关键词
海洋环境数据
平滑支持向量机
时间序列预测
多尺度
marine environment data
smooth support vector machine
time series predictions
multiple time scales
