Counterfactual Maximum Likelihood Estimation for Training Deep Networks

Xinyi Wang, Wenhu Chen, Michael Saxon, William Yang Wang

Keywords: causality, spurious correlation, deep learning, data augmentation, out-of-distribution generalization, natural language processing

TL;DR: We propose a causality-based training framework for deep learning models to reduce the spurious correlations caused by observable confounders, with theoretical analysis as well as experiments on both simulated and real-world datasets.

Abstract: Although deep learning models have driven state-of-the-art performance on a wide array of tasks, they are prone to spurious correlations that should not be learned as predictive clues. To mitigate this problem, we propose a causality-based training framework to reduce the spurious correlations caused by observed confounders. We give theoretical analysis on the underlying general Structural Causal Model (SCM) and propose to perform Maximum Likelihood Estimation (MLE) on the interventional distribution instead of the observational distribution, namely Counterfactual Maximum Likelihood Estimation (CMLE). As the interventional distribution, in general, is hidden from the observational data, we then derive two different upper bounds of the expected negative log-likelihood and propose two general algorithms, Implicit CMLE and Explicit CMLE, for causal predictions of deep learning models using observational data. We conduct experiments on both simulated data and two real-world tasks: Natural Language Inference (NLI) and Image Captioning. The results show that CMLE methods outperform the regular MLE method in terms of out-of-domain generalization performance and reducing spurious correlations, while maintaining comparable performance on the regular evaluations.

Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.

Code: https://github.com/WANGXinyiLinda/CMLE

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2106.03831/code)