Pessimistic Uplift Modeling

TL;DR

This paper introduces Pessimistic Uplift Modeling, a new approach that reduces the impact of data noise on uplift predictions, outperforming existing methods especially in noisy environments.

Contribution

It proposes a novel Pessimistic Uplift Modeling technique that minimizes disturbance effects, improving reliability of treatment effect predictions in noisy data scenarios.

Findings

Outperforms existing uplift methods in high noise environments

Demonstrates robustness of the proposed approach on real and simulated data

Reduces sensitivity of uplift models to data disturbance

Abstract

Uplift modeling is a machine learning technique that aims to model treatment effects heterogeneity. It has been used in business and health sectors to predict the effect of a specific action on a given individual. Despite its advantages, uplift models show high sensitivity to noise and disturbance, which leads to unreliable results. In this paper we show different approaches to address the problem of uplift modeling, we demonstrate how disturbance in data can affect uplift measurement. We propose a new approach, we call it Pessimistic Uplift Modeling, that minimizes disturbance effects. We compared our approach with the existing uplift methods, on simulated and real data-sets. The experiments show that our approach outperforms the existing approaches, especially in the case of high noise data environment.