At last, I both standardize and whiten the independent variables.
Method
The amount of neighborhood is 55. Therefore, the size of my dataset is (55,5), which is inappropriate to use the machine learning algorithm. I used the linear_model.OLS packages in Python (3.6.5) to do the linear regression. From the regression result table, the conclusion is the proportion of different safety hierarchies’ locations do not have a statistically significant relationship with the amount of felony in that neighborhood.