Tools for Uplift Modeling
Uplift barplot
Qini-based feature selection
Univariate quantization
Bivariate quantization
Two-model estimator
Interaction estimator
LASSO path for the penalized logistic regression
Qini curve
Performance of an uplift estimator
Qini curve
Prediction from univariate quantization
Predictions from a two-model estimator
Predictions from an interaction estimator
Qini coefficient
Qini-based uplift regression
Split data with respect to uplift distribution
Uplift barplot for categorical variables
Uplift modeling aims at predicting the causal effect of an action such as a marketing campaign on a particular individual. In order to simplify the task for practitioners in uplift modeling, we propose a combination of tools that can be separated into the following ingredients: i) quantization, ii) visualization, iii) variable selection, iv) parameters estimation and, v) model validation. For more details, see <https://dms.umontreal.ca/~murua/research/UpliftRegression.pdf>.