Variable Selection

Mapping Categorical Predictors to Numeric With Target Encoding

Target encoding is commonly used to map categorical variables to numeric with the objective of facilitating exploratory data analysis and machine learning modeling. This post covers the basics of this method, and explains how and when to use it.

Everything You Don't Need to Know About Variance Inflation Factors

Deep explanation of what Variance Inflation Factors (VIF) are, how they work, what they really mean, and how they are used to manage multicollinearity in linear models.

Multicollinearity Hinders Model Interpretability

In this post, I delve into the intricacies of model interpretation under the influence of multicollinearity, and use R and a toy data set to demonstrate how this phenomenon impacts both linear and machine learning models.