Quantifying Correlations of Machine Learning Models
This addresses safety risks in ML deployments where correlated errors can cause harm, but it is incremental as it quantifies known issues without proposing new solutions.
The paper investigates error correlations between multiple machine learning models in safety-critical applications, finding that aggregated risks are substantial, especially when models share similar algorithms, training datasets, or foundational models.
Machine Learning models are being extensively used in safety critical applications where errors from these models could cause harm to the user. Such risks are amplified when multiple machine learning models, which are deployed concurrently, interact and make errors simultaneously. This paper explores three scenarios where error correlations between multiple models arise, resulting in such aggregated risks. Using real-world data, we simulate these scenarios and quantify the correlations in errors of different models. Our findings indicate that aggregated risks are substantial, particularly when models share similar algorithms, training datasets, or foundational models. Overall, we observe that correlations across models are pervasive and likely to intensify with increased reliance on foundational models and widely used public datasets, highlighting the need for effective mitigation strategies to address these challenges.