Class flipping for uplift modeling and Heterogeneous Treatment Effect estimation on imbalanced RCT data
This work addresses data imbalance issues in causal inference for applications like medical treatments or marketing, offering a practical solution for researchers and practitioners, though it is incremental as it builds on existing transformation-based models.
The paper tackles class and treatment imbalance in uplift modeling and heterogeneous treatment effect estimation from randomized controlled trials by proposing a class-flipping approach that avoids distorting predicted effects and eliminates the need for calibration. Experiments confirm the method's theoretical correctness and show it works for standard classification problems as well.
Uplift modeling and Heterogeneous Treatment Effect (HTE) estimation aim at predicting the causal effect of an action, such as a medical treatment or a marketing campaign on a specific individual. In this paper, we focus on data from Randomized Controlled Experiments which guarantee causal interpretation of the outcomes. Class and treatment imbalance are important problems in uplift modeling/HTE, but classical undersampling or oversampling based approaches are hard to apply in this case since they distort the predicted effect. Calibration methods have been proposed in the past, however, they do not guarantee correct predictions. In this work, we propose an approach alternative to undersampling, based on flipping the class value of selected records. We show that the proposed approach does not distort the predicted effect and does not require calibration. The method is especially useful for models based on class variable transformation (modified outcome models). We address those models separately, designing a transformation scheme which guarantees correct predictions and addresses also the problem of treatment imbalance which is especially important for those models. Experiments fully confirm our theoretical results. Additionally, we demonstrate that our method is a viable alternative also for standard classification problems.