Near-infrared spectrum tea main component detection method based on artificial neural network

The invention discloses a near infrared spectrum tea main component detection method based on an artificial neural network. A cross validation method is used for processing; an error root mean squareis used as an evaluation standard for a correction set and a prediction set; a near infrared spectrum matrix of main components of tea is used as input of the artificial neural network, so that the number of input factors of the artificial neural network is reduced under the condition of ensuring that information is not lost, and the problems of excessive input factors, complex operation and low accuracy of a prediction result when the artificial neural network is used for prediction are effectively solved. 本发明公开了一种基于人工神经网络的近红外光谱茶叶主要成分检测方法，利用交叉验证方法处理方法，对校正集和预测集采用误差均方根作为评估标准，将茶叶主成分的近红外光谱矩阵作为、人工神经网络输入，在保证信息不丢失的情况下，降低了人工神经网络的输入因子个数，有效解决了利用人工神经网络进行预测时，输入因子过多，运算复杂，预测结果准确性不高的问题。