Which technique is most effective for handling skewed numerical features before applying a model that assumes normality?
Data ScienceHard
Data Science — Hard
Which technique is most effective for handling skewed numerical features before applying a model that assumes normality?
Explanation
Box-Cox is a power transformation designed to stabilize variance and make data more closely resemble a normal distribution. Unlike Min-Max or Z-score, which only shift or scale the data without changing the underlying distribution shape, Box-Cox actively transforms the data to correct for skewness, satisfying the normality assumptions required by many linear models.
