Synthetic-Powered Predictive Inference

TL;DR

Synthetic-powered predictive inference (SPI) enhances conformal prediction by integrating synthetic data, leading to more informative prediction sets especially when real data is limited, without sacrificing finite-sample coverage guarantees.

Contribution

The paper introduces SPI, a novel framework that uses synthetic data and a score transporter to improve predictive inference efficiency while maintaining coverage guarantees.

Findings

SPI produces tighter prediction sets with synthetic data.

Experiments show improved efficiency in image classification and tabular regression.

SPI is effective in data-scarce scenarios.

Abstract

Conformal prediction is a framework for predictive inference with a distribution-free, finite-sample guarantee. However, it tends to provide uninformative prediction sets when calibration data are scarce. This paper introduces Synthetic-powered predictive inference (SPI), a novel framework that incorporates synthetic data -- e.g., from a generative model -- to improve sample efficiency. At the core of our method is a score transporter: an empirical quantile mapping that aligns nonconformity scores from trusted, real data with those from synthetic data. By carefully integrating the score transporter into the calibration process, SPI provably achieves finite-sample coverage guarantees without making any assumptions about the real and synthetic data distributions. When the score distributions are well aligned, SPI yields substantially tighter and more informative prediction sets than standard conformal prediction. Experiments on image classification -- augmenting data with synthetic diffusion-model generated images -- and on tabular regression demonstrate notable improvements in predictive efficiency in data-scarce settings.