基于GRNN的酱油种曲孢子数预测模型A Prediction Model of Koji Spores Number in Soy Sauce Production Base on GRNN
张如意,王学雷
摘要(Abstract)：
孢子数是酱油种曲质量的一个重要指标。为优化种曲培养条件,提高种曲质量,文章基于培养过程数据,建立预测种曲孢子数的GRNN神经网络模型,并利用交叉验证确定GRNN模型的最优参数。对比K近邻非参数回归方法、BP神经网络模型,GRNN神经网络模型具有更好的计算稳定性和预测准确性。
关键词(KeyWords)： GRNN;交叉验证;种曲孢子数;数据挖掘
基金项目(Foundation):
作者(Author): 张如意,王学雷
参考文献(References)：
- [1]杨天英.发酵调味品工艺学[M].北京:中国轻工业出版社,2006.
- [2]张国春.浅谈优质种曲的制作要点[J].中国调味品,1998(12):19-21.
- [3]张丽华,何余堂,赵大军,等.影响种曲质量的因素及其控制[J].中国调味品,2006(2):27-29.
- [4]Savita G,Kulkarni,Amit Kumar Chaudhary,SomnathNandi,Modeling and monitoring of batch processes usingprincipal component analysis(PCA)assisted generalized re-gression neural networks(GRNN)[J].Biochemical Engi-neering Journal,2004,18:193-210.
- [5]Jeyamkondan S,Jayas D S,Holley R A.Microbial growth modelling with artificial neural networks[J].International Journal of Food Microbiology,2001,64:343-354.
- [6]Donald F,Specht.A General Regression Neural Network[J].IEEE Transactions on Neural Netwroks,1991,2(6):568-576.
- [7]Sylvain Arlot,Alain Celisse.A survey of cross-validation procedures for model selection[J].Statistics Surveys,2010(4):40-79.
- [8]Ping Zhang.Model Selection Via Multifold Cross Validation[J].The Annals of Statistics,1993,21(1):299-313.
- [9]Eric W M Lee,Chee Peng Lim,Richard K K Yuen,et al.A Hybrid Neural Network Model for Noisy Data Regression[J].IEEE Transactions on Systems,Man,And Cybernet-ics—Part B:Cybernetics,2004,34(2):951-960.