1. Avoid the influence of abnormal data: Sometimes there will be some abnormal data in the data set, which will be greatly separated from the group due to measurement errors or other reasons. If it is not eliminated, it will have a great impact on the research results.
2. Improve the homogeneity of data: Some data belong to different subsets, such as gender, age, etc. If not excluded, it will affect the correlation analysis between variables.
3. Make the data conform to the normal distribution: Many statistical analysis methods are based on the assumption that the data satisfies the normal distribution.