Learning the Distribution Map in Reverse Causal Performative Prediction

TL;DR

This paper introduces a reverse causal modeling approach to learn distribution shifts caused by predictive models affecting agent behavior, with applications in social computing and performative prediction.

Contribution

It proposes a novel reverse causal framework and a statistically justified method to learn distribution shifts driven by agent actions, improving performative prediction.

Findings

Effective in minimizing performative prediction risk

Demonstrates the approach's applicability in social computing scenarios

Provides a microfoundation model for agents' actions

Abstract

In numerous predictive scenarios, the predictive model affects the sampling distribution; for example, job applicants often meticulously craft their resumes to navigate through a screening systems. Such shifts in distribution are particularly prevalent in the realm of social computing, yet, the strategies to learn these shifts from data remain remarkably limited. Inspired by a microeconomic model that adeptly characterizes agents' behavior within labor markets, we introduce a novel approach to learn the distribution shift. Our method is predicated on a reverse causal model, wherein the predictive model instigates a distribution shift exclusively through a finite set of agents' actions. Within this framework, we employ a microfoundation model for the agents' actions and develop a statistically justified methodology to learn the distribution shift map, which we demonstrate to be effective in minimizing the performative prediction risk.