Application of grid search-optimized support vector machine multi-classification parameters in identifying sauce-flavor Baijiu with different processes
To enhance the accuracy of support vector machine(SVM)in the classification and prediction of sauce-flavor(Jiangxiangxing)Baijiu with different processes,a grid search was employed to optimize the parameters of the SVM,and SVM classification prediction model with the optimal pa-rameters was established.Through quantitative analysis of the objective structural characteristics of sauce-flavor Baijiu with different processes,the extracted feature information data was preprocessed(including outlier handling and normalization)and stored as a sample dataset.The sample data were divided into training samples and testing samples.The training samples were used to train the SVM Baijiu brand classification prediction model with optimal parameters,and the test samples were used to predict the classification of the testing samples.The experimental verification demonstrated that the classification recognition rate of the model for different processing technologies reached 94.44％.Compared with the traditional classification algorithms such as SVM,the model could quickly and effectively classify and recognize sauce-flavor Baijiu with different processes,and significantly improving classification accuracy.The improved method was also relatively simple in implementation.
sauce-flavor Baijiu with different processessupport vector machinegrid searchclassification prediction
NETL
NSTL
万方数据
