Datasets › PhysioNet Challenge 2012

PhysioNet Challenge 2012

archive 2025-07-28

The PhysioNet Challenge 2012 dataset is publicly available and contains the de-identified records of 8000 patients in Intensive Care Units (ICU). Each record consists of roughly 48 hours of multivariate time series data with up to 37 features recorded at various times from the patients during their stay such as respiratory rate, glucose etc.

Source: Multi-resolution Networks For Flexible Irregular Time Series Modeling (Multi-FIT) Image Source: https://physionet.org/content/challenge-2016/1.0.0/

Benchmarks archive 2025-07-28

All 5 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

First row (archive order)PaperCode
Time Series Classification PhysioNet Challenge 2012 GRU-D AUC 86.99% Set Functions for Time Series WenjieDu/PyPOTS +1 28 Compare
Multivariate Time Series Imputation PhysioNet Challenge 2012 SAITS MAE (10% of data as GT) 0.186 SAITS: Self-Attention-based Imputation for Time Series WenjieDu/PyPOTS +2 9 Compare
Time Series Analysis PhysioNet Challenge 2012 naive classifier F1 87.47 As easy as APC: overcoming missing data and class... fiorella-wever/APC 7 Compare
Multivariate Time Series Forecasting PhysioNet Challenge 2012 Latent ODE + Poisson mse (10^-3) 2.208 Latent ODEs for Irregularly-Sampled Time Series YuliaRubanova/latent_ode +10 4 Compare
Imputation PhysioNet Challenge 2012 GP-VAE (B-NLST) AUROC 0.743 Seq2Tens: An Efficient Representation of Sequences by... tgcsaba/seq2tens 1 Compare

Papers archive 2025-07-28

14 shown of 14 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 23. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

DateSamples run Syntology
SAITS: Self-Attention-based Imputation for Time Series 3 1 17 Feb 2022 ran 3 of 10 samples (7 unverified)
Self-Supervised Transformer for Sparse and Irregularly Sampled Multivariate Clinical Time-Series 1 1 29 Jul 2021 ran 2 of 2 samples (0 unverified)
As easy as APC: overcoming missing data and class imbalance in time series with self-supervised learning 1 20 29 Jun 2021 not harvested
Multi-Time Attention Networks for Irregularly Sampled Time Series 2 2 25 Jan 2021 ran 2 of 3 samples (1 unverified)
Seq2Tens: An Efficient Representation of Sequences by Low-Rank Tensor Projections 1 1 12 Jun 2020 not harvested
Set Functions for Time Series 2 6 26 Sep 2019 ran 1 of 1 samples (0 unverified)
Interpolation-Prediction Networks for Irregularly Sampled Time Series 1 1 13 Sep 2019 ran 0 of 3 samples (3 unverified)
Latent ODEs for Irregularly-Sampled Time Series 11 7 8 Jul 2019 ran 4 of 8 samples (4 unverified; 3 pointer-only for licence)
Neural Ordinary Differential Equations 56 4 19 Jun 2018 ran 89 of 124 samples (35 unverified; 41 pointer-only for licence)
BRITS: Bidirectional Recurrent Imputation for Time Series 5 1 27 May 2018 ran 2 of 4 samples (2 unverified)
Estimating Missing Data in Temporal Data Streams Using Multi-directional Recurrent Neural Networks 3 1 23 Nov 2017 not harvested
imputeTS: Time Series Missing Value Imputation in R 1 1 1 Jun 2017 not harvested
Recurrent Neural Networks for Multivariate Time Series with Missing Values 7 1 6 Jun 2016 ran 1 of 3 samples (2 unverified; 2 pointer-only for licence)
Multiple imputation using chained equations: issues and guidance for practice 1 1 30 Nov 2010 not harvested

Dataset loaders archive 2025-07-28

1 loader as listed in the archive; links are outbound and not re-checked here.

Tasks archive 2025-07-28

License archive 2025-07-28

No licence recorded in the archive. Absence here is not a statement about the dataset's terms.

Modalities archive 2025-07-28

Languages archive 2025-07-28

No language tagged.

Variants archive 2025-07-28

  • PhysioNet Challenge 2012

1 variant name, as the archive lists them.

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