Papers › As easy as APC: overcoming missing data and class imbalance in time series with...

As easy as APC: overcoming missing data and class imbalance in time series with self-supervised learning

29 Jun 2021arXiv:2106.15577archive 2025-07-28

Fiorella Wever, T. Anderson Keller, Laura Symul, Victor Garcia

High levels of missing data and strong class imbalance are ubiquitous challenges that are often presented simultaneously in real-world time series data. Existing methods approach these problems separately, frequently making significant assumptions about the underlying data generation process in order to lessen the impact of missing information. In this work, we instead demonstrate how a general self-supervised training method, namely Autoregressive Predictive Coding (APC), can be leveraged to overcome both missing data and class imbalance simultaneously without strong assumptions. Specifically, on a synthetic dataset, we show that standard baselines are substantially improved upon through the use of APC, yielding the greatest gains in the combined setting of high missingness and severe class imbalance. We further apply APC on two real-world medical time-series datasets, and show that APC improves the classification performance in all settings, ultimately achieving state-of-the-art AUPRC results on the Physionet benchmark.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

fiorella-wever/APC officialmentioned in papermentioned on GitHubtf report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Self-Supervised LearningTime SeriesTime Series AnalysisTime Series Classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Time Series Analysis PhysioNet Challenge 2012 naive classifier F1 87.47 #1 of 7 Archive leaderboard report
Time Series Analysis PhysioNet Challenge 2012 GRU-D - APC (n = 1) F1 27.3 #2 of 7 Archive leaderboard report
Time Series Analysis PhysioNet Challenge 2012 GRU-APC (n = 1) F1 25.7 #3 of 7 Archive leaderboard report
Time Series Analysis PhysioNet Challenge 2012 GRU-D F1 22.5 #4 of 7 Archive leaderboard report
Time Series Analysis PhysioNet Challenge 2012 GRU F1 22.3 #5 of 7 Archive leaderboard report
Time Series Analysis PhysioNet Challenge 2012 GRU-Simple F1 22.2 #6 of 7 Archive leaderboard report
Time Series Analysis PhysioNet Challenge 2012 GRU-Mean F1 22.1 #7 of 7 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-D - APC (n = 1) AUPRC 55.1 #16 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-Simple AUPRC 53.8 #17 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-D [12] AUPRC 53.7 #18 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-D [12] AUROC 0.863 #18 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-APC (n = 1) AUPRC 53.5 #19 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-D - APC (n = 0) AUPRC 53.3 #20 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-D AUPRC 53.1 #21 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-Forward AUPRC 52 #22 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU AUPRC 51.4 #23 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-APC (n = 0) AUPRC 50.4 #24 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-Mean AUPRC 50.3 #25 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 BRITS [4] AUROC 0.85 #26 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-D [6] AUROC 0.8424 #27 of 28 Archive leaderboard report
Time Series Classification PhysioNet Challenge 2012 GRU-D [4] AUROC 0.834 #28 of 28 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

GRU

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections