2.2.2 Data processing
The data set contains a total of 260 samples, and the relevant data is highly discrete, while data standardization (convert the mean of the sample to 0 and the mean to 1) has the advantages of making the data distribution more concentrated, eliminating the impact of dimensional differences between data, and increasing the interpretability of the machine learning model(De Livera et al., 2015). Therefore, data normalization was performed on our dataset to improve the data structure.