Mitigating these biases relies on exposing and removingnuisance components.

For all considered pseudo-outcomes it holds thatEP [ Yη|X = x] = τ(x) - they are unbiased for CATEwhenηisknown. The testable implications oftheshared structure bias, asencoded byhyperparameters suchasλ2,however,aredifferentfor(i)thePOestimationand(ii) 2 the CATE estimation problems, which is a feature that we would suggest to exploit in choosing hyperparameter settings. Unfortunately,good performance on estimation of the POs is not sufficient. Illustrative results In Figure 1 we present illustrative results on Setup B withn0 = n1 = 2000, and observe that following our heuristic of increasingλ2 until factual performance deteriorates would almost always lead to choosing the best hyperparameter setting; for both hard and flexible approach this suggests a switch fromλ = 10 1 to λ2 = 10 2 as ρ increases1. Therefore, FlexTENet also generalizes the SNet class discussed in [4], which includes PO-specific feature spaces3.

At this point, a range of sophisticated solutions exist which reduce the effect of confounding by balancing the covariate space [3, 4], importance weighting [11, 12, 13, 14] or propensity drop-out [15].

This study proposes an end-to-end algorithm for policy learning in causal inference. We observe data consisting of covariates, treatment assignments, and outcomes, where only the outcome corresponding to the assigned treatment is observed. The goal of policy learning is to train a policy from the observed data, where a policy is a function that recommends an optimal treatment for each individual, to maximize the policy value. In this study, we first show that maximizing the policy value is equivalent to minimizing the mean squared error for the conditional average treatment effect (CATE) under $\{-1, 1\}$ restricted regression models. Based on this finding, we modify the causal forest, an end-to-end CATE estimation algorithm, for policy learning. We refer to our algorithm as the causal-policy forest. Our algorithm has three advantages. First, it is a simple modification of an existing, widely used CATE estimation method, therefore, it helps bridge the gap between policy learning and CATE estimation in practice. Second, while existing studies typically estimate nuisance parameters for policy learning as a separate task, our algorithm trains the policy in a more end-to-end manner. Third, as in standard decision trees and random forests, we train the models efficiently, avoiding computational intractability.

Firms often develop targeting policies to personalize marketing actions and improve incremental profits. Effective targeting depends on accurately separating customers with positive versus negative treatment effects. We propose an approach to estimate the conditional average treatment effects (CATEs) of marketing actions that aligns their estimation with the firm's profit objective. The method recognizes that, for many customers, treatment effects are so extreme that additional accuracy is unlikely to change the recommended actions. However, accuracy matters near the decision boundary, as small errors can alter targeting decisions. By modifying the firm's objective function in the standard profit maximization problem, our method yields a near-optimal targeting policy while simultaneously estimating CATEs. This introduces a new perspective on CATE estimation, reframing it as a problem of profit optimization rather than prediction accuracy. We establish the theoretical properties of the proposed method and demonstrate its performance and trade-offs using synthetic data.