DANCE: Doubly Adaptive Neighborhood Conformal Estimation
This addresses uncertainty quantification for pre-trained deep learning models in classification tasks, representing an incremental improvement over existing conformal methods.
The paper tackles the problem of inefficient, overly conservative prediction sets in conformal prediction for pre-trained models by proposing DANCE, a doubly locally adaptive nearest-neighbor based conformal algorithm that demonstrates superior set size efficiency and robustness across various datasets compared to state-of-the-art baselines.
The recent developments of complex deep learning models have led to unprecedented ability to accurately predict across multiple data representation types. Conformal prediction for uncertainty quantification of these models has risen in popularity, providing adaptive, statistically-valid prediction sets. For classification tasks, conformal methods have typically focused on utilizing logit scores. For pre-trained models, however, this can result in inefficient, overly conservative set sizes when not calibrated towards the target task. We propose DANCE, a doubly locally adaptive nearest-neighbor based conformal algorithm combining two novel nonconformity scores directly using the data's embedded representation. DANCE first fits a task-adaptive kernel regression model from the embedding layer before using the learned kernel space to produce the final prediction sets for uncertainty quantification. We test against state-of-the-art local, task-adapted and zero-shot conformal baselines, demonstrating DANCE's superior blend of set size efficiency and robustness across various datasets.