NormForm

Value parameters

robust

whether to add a small value (ε) to the standard deviation to avoid DBZ when a whole column has zero stdev (σ) it should be removed in pre-processing

x_

the input vector or matrix to be transformed (needed to get w)

Attributes

See also

www.geeksforgeeks.org/machine-learning/standardscaler-minmaxscaler-and-robustscaler-techniques-ml/ x -> (x - μ)/σ robust version: (x - μ)/(σ + ε) or (x - μ)/√(σ² + ε)

Graph

Reset zoomHide graphShow graph

Supertypes

trait Transform

class Object

trait Matchable

class Any