Computer scienceData scienceInstrumentsScikit-learnData preprocessing with scikit-learn

Additional feature scaling techniques in scikit-learn

The power transformation difference

Report a typo

What is the main difference between the Box-Cox and the Yeo-Johnson transformations?

Select one option from the list

Yeo-Johnson is suatable for both the negative and the positive values, while Box-Cox requires the values to be strictly positive

Box-Cox makes the data's distribution normal, while Yeo-Johnson makes the distribution uniform

With all other parameters set to default, the Box-Cox transformation does not apply zero mean and unit variance, and Yeo-Johnson applies them by default

Yeo-Johnson outperforms the Box-Cox transformation in supressing the outliers

___

Create a free account to access the full topic