Predictive Inference With Fast Feature Conformal Prediction
This work addresses a practical bottleneck for researchers and practitioners using conformal prediction in deep learning by making it faster, though it is incremental as it builds directly on FCP.
The paper tackles the computational inefficiency of Feature Conformal Prediction (FCP) in uncertainty quantification by introducing Fast Feature Conformal Prediction (FFCP), which uses a novel non-conformity score and Taylor expansion to approximate non-linear operations, achieving comparable performance to FCP while reducing computational time by approximately 50x.
Conformal prediction is widely adopted in uncertainty quantification, due to its post-hoc, distribution-free, and model-agnostic properties. In the realm of modern deep learning, researchers have proposed Feature Conformal Prediction (FCP), which deploys conformal prediction in a feature space, yielding reduced band lengths. However, the practical utility of FCP is limited due to the time-consuming non-linear operations required to transform confidence bands from feature space to output space. In this paper, we introduce Fast Feature Conformal Prediction (FFCP), which features a novel non-conformity score and is convenient for practical applications. FFCP serves as a fast version of FCP, in that it equivalently employs a Taylor expansion to approximate the aforementioned non-linear operations in FCP. Empirical validations showcase that FFCP performs comparably with FCP (both outperforming the vanilla version) while achieving a significant reduction in computational time by approximately 50x. The code is available at https://github.com/ElvisWang1111/FastFeatureCP