【摘要】 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。更多还原

【Abstract】 The spores number is an important index of koji quality.In order to optimize cultivation conditions and improve the quality of soy sauce koji,Based on process data,from the perspective of data mining,a General Regression Neural Network(GRNN) with cross validation for optimizing model parameter was proposed to predict the final spores number.Compared with K-nearest neighbor method and back propagation neural network,GRNN shows a better performance.更多还原