Explainable uplift modeling via linearized kernel feature maps, providing a collection of meta-learners.
Install using pip:
pip install xuplift- Regressor: High-performance regression engine for outcome and residual modeling.
- Classifier: Optimized binary classifier for precise propensity score estimation.
- RLearner: Advanced residual-on-residual estimator with built-in 2-fold cross-fitting to ensure unbiased treatment effect estimation.
- XLearner: Optimized cross-learner designed to handle significantly unbalanced treatment groups.
- TLearner/SLearner: Standard two-model and single-model estimators for baseline causal analysis.