The Importance of Time in Causal Algorithmic Recourse

TL;DR

This work motivates the need to integrate the temporal dimension into causal algorithmic recourse methods to enhance recommendations' plausibility and reliability and highlights the significance of the role of time.

Abstract

The application of Algorithmic Recourse in decision-making is a promising field that offers practical solutions to reverse unfavorable decisions. However, the inability of these methods to consider potential dependencies among variables poses a significant challenge due to the assumption of feature independence. Recent advancements have incorporated knowledge of causal dependencies, thereby enhancing the quality of the recommended recourse actions. Despite these improvements, the inability to incorporate the temporal dimension remains a significant limitation of these approaches. This is particularly problematic as identifying and addressing the root causes of undesired outcomes requires understanding time-dependent relationships between variables. In this work, we motivate the need to integrate the temporal dimension into causal algorithmic recourse methods to enhance recommendations' plausibility and reliability. The experimental evaluation highlights the significance of the role of time in this field.

Figures (6)

• Figure 1: A causal graph illustrating the relationship between college education, individual skill, and job salary discussed in glymour2016causal.
• Figure 2: An example of a causal graph and a simple way to incorporate time information over it. A wavy edge weight $\tau_{ij}$ is meant to represent time between intervention over node $i$ and the observed effect on node $j$.
• Figure 3: The German Credit inspired causal DAG.
• Figure 4: Actionable weighted DAG where the coefficients represent the response times for parent-child relationships.
• Figure 5: Example of possibly different costs based on the scale of distributions considered.