摘要
鱼类市场价格是影响渔民收益的重要因素,因此如何能够准确分析和预测鱼类的价格是一个十分重要的问题。论文在考虑了BP网络原形的一些缺点和不足,尝试在传统BP算法的基础上将网络神经元中的激励函数换成小波子函数,组建成的小波神经网络。通过对鲈鱼价格的预测,验证了小波神经网络的可行性,然后基于阿里巴巴的鲈鱼价格的历史数据,验证了该方法的合理性。最后开展对新疆乌伦古湖水产综合基地三类主要鱼类价格预测的实证研究,研究结果表明河鲈和梭鲈的价格会有小幅的波动,但高白鲑的价格会有大幅的提升。因此,小波神经网络作为一种传统神经网络的改进,可用于预测鱼类价格的短期预测,为水产基地的决策提供依据。
The market price of fish is an important element effecting the earnings of fishermen. Hence, it is an interesting issue to study the accurate analysis and prediction of the price of the fish. This paper considers the shortcoming of traditional BP net algorithm and changes the activation function with wavelet function, which forms a new wavelet neural network. This model is firstly applied in the forecast of striped bass with the data from Alibaba indicating the feasibility of this method. Then the data of three different fishes in aquatic integrated base in Wulungu Lake, Xinjiang, are employed for the application of this model, which can provide the basis for strategies. The result shows that the bass will have a little price fall at first and then a rising price, the price of zander will decrease slow and coregonus peled's price will go up increasingly. Therefore, the wavelet neural network, as a kind of improvement of traditional which can provide the basis for strategies. network, is applicable in the price forecast of fish,
出处
《中国渔业经济》
2014年第4期61-66,共6页
Chinese Fisheries Economics
关键词
神经网络
小波函数
鱼类价格
预测
neural networks
wavelet function
fish price
forecasting