202406071418
Status: #idea
Tags: Transformations
State: #nascient

Transformations

Transformations to fix heteroschedascity

A common reason for the violation of the assumption of homoschedascity is that the fact that the distribution of $Y$ depends on the mean, making the variance depend on the $E(X)$ which is problematic.

In this note, we explore how to fix that.
Variance Stabilizing Transformations

Transformation to linearize the model

What do we do if the plot of $x$ and $y$ is not linear?

Linearizing Transformations

References