Learning Feasible Causal Algorithmic Recourse: A Prior Structural Knowledge Free Approach

Algorithmic recourse (AR) has made significant progress by identifying small perturbations in input features that can alter predictions, which provide a data-centric approach to understand decisions from diverse black-box models on the Web. Towards the feasibility issue, i.e., whether the recoursed examples provides actionable and reliable recommendations to end-users, causal algorithmic recourse have incorporated structural causal model (SCM) to preserve the realistic constraints among input features. For instance, preserving structural causal knowledge between "age" and "educational level" can avoid generating samples with decreasing age and increasing educational level. However, previous causal AR methods suffer from the requirement of prior structural causal knowledge, e.g., prior causal graph or the whole SCM, which restricts the realistic application of causal AR methods. To bridge this gap, we aim to develop a novel framework for causal algorithmic recourse that does not rely on neither prior causal graph or prior SCM. Since identifying counterfactuals without causal graph is impossible, we instead propose to approximate and constrain the variation of the perturbed components, i.e., the exogenous noise variables, by formulating the generation of AR as the structure-preserving intervention. With the aid of development in non-linear Independent Component Analysis (ICA), our method can further achieve theoretically guaranteed constraints on such variation of exogeneous variables. Experimental results on synthetic, semi-synthetic, and real-world data demonstrate the effectiveness of our proposed methods without any prior causal graph or SCM knowledge.