CAJ | 학술논문

蔬菜价格波动性大、季节性强，受供求关系、季节、气候等多种因素的影响，蔬菜价格具有高度的非线性、高噪声的属性加大了预测的难度。因此运用基于传统的MSE准则函数应用于蔬菜价格预测的建模，以样本的实际输出和神经网络的预测输出的平方差为误差准则函数，来进行人工神经网络参数学习，通过MSE准则的TSK模糊系统模型来预测蔬菜的价格，最后通过实例验证此方法科学有效。
소채개격파동성대、계절성강，수공구관계、계절、기후등다충인소적영향，소채개격구유고도적비선성、고조성적속성가대료예측적난도。인차운용기우전통적MSE준칙함수응용우소채개격예측적건모，이양본적실제수출화신경망락적예측수출적평방차위오차준칙함수，래진행인공신경망락삼수학습，통과MSE준칙적TSK모호계통모형래예측소채적개격，최후통과실례험증차방법과학유효。
Vegetable price fluctuates greatly and seasonally. It is influenced by the supply and demand relationship, season, and climate. Vegetable price is difficult to forecast for its high degree of non-linear and high-noise properties. For this reason, the paper applies the traditional MSE guidelines function to vegetable price forecast modeling, makes artificial neural network parameters studying by the Squared-Error of the actual output and the neural network forecast output, forecast vegetable price based on the Mse guidelines TSK fuzzy system, and finally proves that the method is efficient by way of example.