METHOD AND APPARATUS FOR ESTIMATING TREATMENT EFFECT USING PROPENSITY SCORE RESIDUAL IN LEAST SQUARES REGRESSION ANALYSIS

摘要

The first residual term calculating unit calculates the first parameter value and the propensity score based on the processing variable model for the processing variable and the distribution model for the first error term included in the processing variable model, and subtracts the propensity score from the processing variable. Calculating a first residual term, wherein the second residual term calculating unit calculates at least one second parameter value included in an approximation model for the conditional average value of the result variable based on the first parameter value and applies the approximation model to the approximation model Generating a model, calculating a second residual term obtained by subtracting the parameter application approximation model from the result variable, and the processing effect estimating unit estimates the processing effect parameter of the processing variable on the result variable based on the first residual term and the second residual term. The present invention relates to a method and apparatus for estimating a processing effect using a propensity score residual term in a least squares regression analysis, comprising: Will.