摘要 以深度学习为基础,结合Tensorflow与Keras框架,建立了基于深度学习的白酒品牌分类预测模型。通过电子舌(阵列式传感器)对待测白酒的特征信息进行采集,并与已知的待测白酒样品类别结合建立测试样本数据集,通过训练集与测试集对基于深度学习的白酒品牌分类预测模型进行训练与性能检验。结果表明,该预测模型的白酒品牌识别率达99.987%,准确率较高。 A deep learning-based model for the classification and prediction of Baijiu wine brands is developed,combining with the Tensorflow and Keras frameworks.The test sample data set is established by collecting the feature information of Baijiu wine to be tested through an electronic tongue(an array sensor),combining the known Baijiu wine categories with that of the unknow one(to be test).Then the deep learning-based Baijiu wine brand classification prediction model was trained and tested for,preparing for the performance by using the trainedand tested sets.The results showed that the prediction model achieves 99.987%of Baijiu wine brand recognition rate,showing a high accuracy.
出处 《食品与机械》 北大核心 2021年第4期68-71,79,共5页 Food and Machinery
基金 四川省科技计划项目(编号:2021YFS0339) 四川省科技成果转移转化示范项目(编号:2020ZHCG0040) 四川省重大科技专项项目(编号:2018GZDZX0045)。
关键词 深度学习 分类预测 电子舌 实用性 高效性 deep learning classification prediction electronic tongue practicality high efficiency
分类号 F42 [经济管理—产业经济]
