基于先验神经网络的港口物流量对海洋经济增长预测——以浙江省为例

Analysis on the Relationship between the Amount of Port Logistics to Coastal Marine Economic Growth and Gorecasting Based on Priori Neural Network:Case of Zhejiang Province

提出了一种基于先验知识神经网络预测模型,分别采用时间序列和多变量建模方法对反映浙江省海洋经济发展水平的各指标进行建模预测,与其他建模方法(如GM(1,1)时间序列模型)相比,有更好的综合性能。最后采用PKNN时间序列模型对2011-2020年浙江省沿海地区海洋经济发展水平进行预测,结果表明预测得到的2015年浙江海洋经济产值接近浙江省规划值。

On the basis of qualitative analysis on the relationship between the amount of port logistics to coastal marine economic growth in Zhejiang Province, the paper points out a priori knowledge-based new neural network model. Multivariate and time series modeling method are respectively applied to predict indexes reflecting the marine economy development in Zhejiang Province. Compared to other modeling methods （such as GM （1,1）time series model）, it has a better overall performance. Finally, we use PKNN time series model to predict the level of economic development in Zhejiang coastal ocean from 2011 to 2020, and the results show that marine economy prediction in 2015 near the Zhejiaug Province planning value, that provide important theory and practice cornerstone to the construction of marine economy in Zhejiang province.