IPW

A classical method for estimating average treatment effects in causal inference. It constructs weighted averages of the treatment and control groups using the inverse of the propensity score as weights. Its unbiasedness relies on the correct specification of the propensity score model.