Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification

About

Black-box machine learning models are now routinely used in high-risk settings, like medical diagnostics, which demand uncertainty quantification to avoid consequential model failures. Conformal prediction is a user-friendly paradigm for creating statistically rigorous uncertainty sets/intervals for the predictions of such models. Critically, the sets are valid in a distribution-free sense: they possess explicit, non-asymptotic guarantees even without distributional assumptions or model assumptions. One can use conformal prediction with any pre-trained model, such as a neural network, to produce sets that are guaranteed to contain the ground truth with a user-specified probability, such as 90%. It is easy-to-understand, easy-to-use, and general, applying naturally to problems arising in the fields of computer vision, natural language processing, deep reinforcement learning, and so on. This hands-on introduction is aimed to provide the reader a working understanding of conformal prediction and related distribution-free uncertainty quantification techniques with one self-contained document. We lead the reader through practical theory for and examples of conformal prediction and describe its extensions to complex machine learning tasks involving structured outputs, distribution shift, time-series, outliers, models that abstain, and more. Throughout, there are many explanatory illustrations, examples, and code samples in Python. With each code sample comes a Jupyter notebook implementing the method on a real-data example; the notebooks can be accessed and easily run using our codebase.

Anastasios N. Angelopoulos, Stephen Bates• 2021

Related benchmarks

TaskDatasetResultRank
ClassificationCase Study 2
F1 Score8.6
50
Error AttributionGSM8k Left Dense
Removal Rate82
30
Error AttributionGSM8k Mid Dense
Removal Rate75
30
Error AttributionMATH
Removal Rate62
30
Error AttributionGSM8k Right Dense
Removal Rate63
30
Error AttributionWho&When--
30
Uncertainty QuantificationMSD Task01 (test)
Coverage (%)94.22
24
ClassificationImageNet
WUC0.023
24
ClassificationImageNet V2
WUC0.043
24
Image ClassificationImageNet V2
Coverage (Cov)71
24
Showing 10 of 26 rows

Other info

Follow for update