Current browse context:
stat.ML
Change to browse by:
References & Citations
Statistics > Machine Learning
Title: Conformal Prediction Intervals with Temporal Dependence
(Submitted on 25 May 2022 (v1), last revised 2 Oct 2022 (this version, v3))
Abstract: Cross-sectional prediction is common in many domains such as healthcare, including forecasting tasks using electronic health records, where different patients form a cross-section. We focus on the task of constructing valid prediction intervals (PIs) in time series regression with a cross-section. A prediction interval is considered valid if it covers the true response with (a pre-specified) high probability. We first distinguish between two notions of validity in such a setting: cross-sectional and longitudinal. Cross-sectional validity is concerned with validity across the cross-section of the time series data, while longitudinal validity accounts for the temporal dimension. Coverage guarantees along both these dimensions are ideally desirable; however, we show that distribution-free longitudinal validity is theoretically impossible. Despite this limitation, we propose Conformal Prediction with Temporal Dependence (CPTD), a procedure that is able to maintain strict cross-sectional validity while improving longitudinal coverage. CPTD is post-hoc and light-weight, and can easily be used in conjunction with any prediction model as long as a calibration set is available. We focus on neural networks due to their ability to model complicated data such as diagnosis codes for time series regression, and perform extensive experimental validation to verify the efficacy of our approach. We find that CPTD outperforms baselines on a variety of datasets by improving longitudinal coverage and often providing more efficient (narrower) PIs.
Submission history
From: Zhen Lin [view email][v1] Wed, 25 May 2022 17:45:04 GMT (153kb,D)
[v2] Thu, 26 May 2022 02:57:17 GMT (159kb,D)
[v3] Sun, 2 Oct 2022 20:59:36 GMT (167kb,D)
Link back to: arXiv, form interface, contact.