A Bayesian Non-Stationary Heteroskedastic Time Series Model for Multivariate Critical Care Data

We propose a multivariate GARCH model for non-stationary health time series by modifying the variance of the observations of the standard state space model. The proposed model provides an intuitive way of dealing with heteroskedastic data using the conditional nature of state space models. We follow the Bayesian paradigm to perform the inference procedure. In particular, we use Markov chain Monte Carlo methods to obtain samples from the resultant posterior distribution. Due to the natural temporal correlation structure induced on model parameters, we use the forward filtering backward sampling algorithm to efficiently obtain samples from the posterior distribution. The proposed model also handles missing data in a fully Bayesian fashion. We validate our model on synthetic data, and then use it to analyze a data set obtained from an intensive care unit in a Montreal hospital. We further show that our proposed models offer better performance, in terms of WAIC, than standard state space models. The proposed model provides a new way to model multivariate heteroskedastic non-stationary time series data and the simplicity in applying the WAIC allows us to compare competing models.

[1]  Shenda Hong,et al.  A Review of Deep Learning Methods for Irregularly Sampled Medical Time Series Data , 2020, ArXiv.

[2]  Jonathan A. C. Sterne,et al.  Use of machine learning to analyse routinely collected intensive care unit data: a systematic review , 2019, Critical Care.

[3]  Ke-Jia Chen,et al.  Sequential online prediction in the presence of outliers and change points: an instant temporal structure learning approach , 2019, Neurocomputing.

[4]  Concha Bielza,et al.  A Fast Metropolis-Hastings Method for Generating Random Correlation Matrices , 2018, IDEAL.

[5]  Roger G. Mark,et al.  Reproducibility in critical care: a mortality prediction case study , 2017, MLHC.

[6]  Diane J. Cook,et al.  A survey of methods for time series change point detection , 2017, Knowledge and Information Systems.

[7]  Yan Liu,et al.  Recurrent Neural Networks for Multivariate Time Series with Missing Values , 2016, Scientific Reports.

[8]  Peter Szolovits,et al.  MIMIC-III, a freely accessible critical care database , 2016, Scientific Data.

[9]  Eric Ruggieri,et al.  An exact approach to Bayesian sequential change point detection , 2016, Comput. Stat. Data Anal..

[10]  Peter Szolovits,et al.  Predicting ICU Mortality Risk by Grouping Temporal Trends from a Multivariate Panel of Physiologic Measurements , 2016, AAAI.

[11]  Shamim Nemati,et al.  Machine Learning and Decision Support in Critical Care , 2016, Proceedings of the IEEE.

[12]  Ram Akella,et al.  Dynamically Modeling Patient's Health State from Electronic Medical Records: A Time Series Approach , 2015, KDD.

[13]  Peter Szolovits,et al.  A Multivariate Timeseries Modeling Approach to Severity of Illness Assessment and Forecasting in ICU with Sparse, Heterogeneous Clinical Data , 2015, AAAI.

[14]  Phyllis K Stein,et al.  Challenges of heart rate variability research in the ICU. , 2013, Critical care medicine.

[15]  Shahram Ebadollahi,et al.  Real-time prognosis of ICU physiological data streams , 2010, 2010 Annual International Conference of the IEEE Engineering in Medicine and Biology.

[16]  Steven Reece,et al.  Sequential Bayesian Prediction in the Presence of Changepoints and Faults , 2010, Comput. J..

[17]  John K Kruschke,et al.  Bayesian data analysis. , 2010, Wiley interdisciplinary reviews. Cognitive science.

[18]  Sumio Watanabe,et al.  Asymptotic Equivalence of Bayes Cross Validation and Widely Applicable Information Criterion in Singular Learning Theory , 2010, J. Mach. Learn. Res..

[19]  T. Komatsu,et al.  [Analysis of heart rate variability]. , 2009, Masui. The Japanese journal of anesthesiology.

[20]  A. Porta,et al.  Heart rate variability explored in the frequency domain: A tool to investigate the link between heart and behavior , 2009, Neuroscience & Biobehavioral Reviews.

[21]  Ryan P. Adams,et al.  Bayesian Online Changepoint Detection , 2007, 0710.3742.

[22]  G. Kitagawa,et al.  Information Criteria and Statistical Modeling , 2007 .

[23]  U. Rajendra Acharya,et al.  Heart rate variability: a review , 2006, Medical and Biological Engineering and Computing.

[24]  Sylvia Frühwirth-Schnatter,et al.  Finite Mixture and Markov Switching Models , 2006 .

[25]  Benoit M Dawant,et al.  Reduced Heart Rate Volatility: An Early Predictor of Death in Trauma Patients , 2004, Annals of surgery.

[26]  V. Somers,et al.  Heart Rate Variability: , 2003, Journal of cardiovascular electrophysiology.

[27]  R. Kohn,et al.  On Gibbs sampling for state space models , 1994 .

[28]  S. Frühwirth-Schnatter.  Data Augmentation and Dynamic Linear Models , 1994 .

[29]  T. Bollerslev,et al.  Modelling the Coherence in Short-run Nominal Exchange Rates: A Multivariate Generalized ARCH Model , 1990 .

[30]  S Cerutti,et al.  Spectral and cross-spectral analysis of heart rate and arterial blood pressure variability signals. , 1986, Computers and biomedical research, an international journal.

[31]  T. Bollerslev,et al.  Generalized autoregressive conditional heteroskedasticity , 1986 .

[32]  Thar Baker,et al.  Prospects of Machine and Deep Learning in Analysis of Vital Signs for the Improvement of Healthcare Services , 2020, Nature-Inspired Computation in Data Mining and Machine Learning.

[33]  Shamim Nemati,et al.  A Model-Based Machine Learning Approach to Probing Autonomic Regulation From Nonstationary Vital-Sign Time Series , 2018, IEEE Journal of Biomedical and Health Informatics.

[34]  Ute Beyer,et al.  Bayesian Forecasting And Dynamic Models , 2016 .

[35]  A. Malliani,et al.  Heart rate variability signal processing: a quantitative approach as an aid to diagnosis in cardiovascular pathologies. , 1987, International journal of bio-medical computing.