In-Person Poster presentation / poster accept

Avoiding spurious correlations via logit correction

Sheng Liu · Xu Zhang · Nitesh Sekhar · Yue Wu · Prateek Singhal · Carlos Fernandez-Granda

MH1-2-3-4 #43

Keywords: [ Deep Learning and representational learning ] [ robust learning ] [ spurious correlation ] [ empirical risk minimization ]

[ Abstract ]
[ OpenReview
Mon 1 May 2:30 a.m. PDT — 4:30 a.m. PDT


Empirical studies suggest that machine learning models trained with empirical risk minimization (ERM) often rely on attributes that may be spuriously correlated with the class labels. Such models typically lead to poor performance during inference for data lacking such correlations. In this work, we explicitly consider a situation where potential spurious correlations are present in the majority of training data. In contrast with existing approaches, which use the ERM model outputs to detect the samples without spurious correlations and either heuristically upweight or upsample those samples, we propose the logit correction (LC) loss, a simple yet effective improvement on the softmax cross-entropy loss, to correct the sample logit. We demonstrate that minimizing the LC loss is equivalent to maximizing the group-balanced accuracy, so the proposed LC could mitigate the negative impacts of spurious correlations. Our extensive experimental results further reveal that the proposed LC loss outperforms state-of-the-art solutions on multiple popular benchmarks by a large margin, an average 5.5% absolute improvement, without access to spurious attribute labels. LC is also competitive with oracle methods that make use of the attribute labels.

Chat is not available.