Evidential Conditional Neural Processes

TL;DR

Evidential Conditional Neural Processes (ECNP) enhance CNP models by providing detailed uncertainty quantification and robustness through a hierarchical Bayesian evidential framework, improving few-shot learning performance.

Contribution

ECNP introduces a hierarchical Bayesian evidential structure to CNPs, enabling fine-grained uncertainty decomposition and robustness, with theoretical analysis and extensive empirical validation.

Findings

ECNP achieves better uncertainty quantification in few-shot tasks.

ECNP demonstrates robustness to noisy training data.

Experimental results show improved performance over standard CNPs.

Abstract

The Conditional Neural Process (CNP) family of models offer a promising direction to tackle few-shot problems by achieving better scalability and competitive predictive performance. However, the current CNP models only capture the overall uncertainty for the prediction made on a target data point. They lack a systematic fine-grained quantification on the distinct sources of uncertainty that are essential for model training and decision-making under the few-shot setting. We propose Evidential Conditional Neural Processes (ECNP), which replace the standard Gaussian distribution used by CNP with a much richer hierarchical Bayesian structure through evidential learning to achieve epistemic-aleatoric uncertainty decomposition. The evidential hierarchical structure also leads to a theoretically justified robustness over noisy training tasks. Theoretical analysis on the proposed ECNP establishes the relationship with CNP while offering deeper insights on the roles of the evidential parameters. Extensive experiments conducted on both synthetic and real-world data demonstrate the effectiveness of our proposed model in various few-shot settings.