2016

Modeling Missing Data in Clinical Time Series with RNNs

Lipton, Zachary C., Kale, David C., Wetzel, Randall

Understand

We demonstrate a simple strategy to cope with missing data in sequential inputs, addressing the task of multilabel classification of diagnoses given clinical time series.

Collected from the pediatric intensive care unit (PICU) at Children's Hospital Los Angeles, our data consists of multivariate time series of observations.
The measurements are irregularly spaced, leading to missingness patterns in temporally discretized sequences.
While these artifacts are typically handled by imputation, we achieve superior predictive performance by treating the artifacts as features.

Built on

Applied multiple regression/correlation analysis for the behavioral sciences
Jacob Cohen and Patricia Cohen · 1975
Earlier work this paper cites.
Generalization of backpropagation with application to a recurrent gas market model
Paul J Werbos · 1988
Earlier work this paper cites.
Learning long-term dependencies with gradient descent is difficult
Yoshua Bengio, Patrice Simard, and Paolo Frasconi · 1994
Earlier work this paper cites.
Recurrent neural networks for missing or asynchronous data
Yoshua Bengio and Francois Gingras · 1996
Earlier work this paper cites.
Long short-term memory
Sepp Hochreiter and Jürgen Schmidhuber · 1997
Earlier work this paper cites.
A solution for missing data in recurrent neural networks with an application to blood glucose prediction
Volker Tresp and Thomas Briegel · 1998
Earlier work this paper cites.
Learning to forget: Continual prediction with LSTM
Felix A. Gers, Jürgen Schmidhuber, and Fred Cummins · 2000
Earlier work this paper cites.
Missing data
Paul D Allison · 2001
Earlier work this paper cites.
Linking auditory scene analysis and robust asr by missing data techniques
Jon Barker, Phil Green, and Martin Cooke · 2001
Earlier work this paper cites.
Speech recognition with missing data using recurrent neural nets
Shahla Parveen and P Green · 2001
Earlier work this paper cites.

Similar

A review of methods for missing data
Therese D Pigott · 2001
Cited alongside, same era.
Learning precise timing with LSTM recurrent networks
Felix A. Gers, Nicol N. Schraudolph, and Jürgen Schmidhuber · 2003
Cited alongside, same era.
The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents
National High Blood Pressure Education Program Working Group on Children and Adolescents · 2004
Cited alongside, same era.
International statistical classification of diseases and related health problems
World Health Organization · 2004
Cited alongside, same era.
A fast learning algorithm for deep belief nets
Geoffrey E. Hinton, Simon Osindero, and Yee Whye Teh · 2006
Cited alongside, same era.
Pattern classification with missing data: a review
Pedro J García-Laencina, José-Luis Sancho-Gómez, and Aníbal R Figueiras-Vidal · 2010
Cited alongside, same era.
Normal ranges of heart rate and respiratory rate in children from birth to 18 years: A systematic review of observational studies
Susannah Fleming, Matthew Thompson, Richard Stevens, Carl Heneghan, Annette Plüddemann, Ian Maconochie, Lionel Tarassenko, and David Mant · 2011
Cited alongside, same era.
Unsupervised pattern discovery in electronic health care data using probabilistic clustering models
Ben M. Marlin, David C. Kale, Robinder G. Khemani, and Randall C. Wetzel · 2012
Cited alongside, same era.
Patient risk stratification for hospital-associated c. diff as a time-series classification task
Jenna Wiens, Eric Horvitz, and John V Guttag · 2012
Cited alongside, same era.

Then

Computational phenotype discovery using unsupervised feature learning over noisy, sparse, and irregular clinical data
Thomas A. Lasko, Joshua C. Denny, and Mia A. Levy · 2013
Later among the works it cites.
Optimal thresholding of classifiers to maximize F1 measure
Zachary C Lipton, Charles Elkan, and Balakrishnan Naryanaswamy · 2014
Later among the works it cites.
Intelligible models for healthcare: Predicting pneumonia risk and hospital 30-day readmission
Rich Caruana, Yin Lou, Johannes Gehrke, Paul Koch, Marc Sturm, and Noemie Elhadad · 2015
Later among the works it cites.
Deep computational phenotyping
Zhengping Che, David C. Kale, Wenzhe Li, Mohammad Taha Bahadori, and Yan Liu · 2015
Later among the works it cites.
A targeted real-time early warning score (trewscore) for septic shock
Katharine E Henry, David N Hager, Peter J Pronovost, and Suchi Saria · 2015
Later among the works it cites.
A critical review of recurrent neural networks for sequence learning
Original
Zachary C. Lipton, John Berkowitz, and Charles Elkan · 2015
Later among the works it cites.
The digital revolution in phenotyping
Anika Oellrich et al · 2015
Later among the works it cites.
The mythos of model interpretability
Zachary C Lipton · 2016
Closest in time.
Learning to diagnose with lstm recurrent neural networks
Zachary C Lipton, David C Kale, Charles Elkan, and Randall Wetzell · 2016
Closest in time.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…

Reading the bibliography…