Multitask Gaussian Processes for Multivariate Physiological Time-Series Analysis

By Robert Dürichen, Marco A. F. Pimentel, Lei A. Clifton, A. Schweikard, David A. Clifton. 2015

167 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Biomedical Engineering, Surgery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Biomedical Engineering

In Surgery

Citations in the corpus

A Reinforcement Learning Approach to Weaning of Mechanical Ventilation in Intensive Care Units (2017, methodology): “One commonly used approach to resolve missing and irregularly sampled time series data is Gaussian processes (GPs, [Stegle et al., 2008, Dürichen et al., 2015, Ghassemi et al., 2015]).”
Learning to Detect Sepsis with a Multitask Gaussian Process RNN Classifier (2017, methodology): “, 2015) and (Durichen et al., 2015) used a similar model to ours, but instead focused more on forecasting of vitals to predict clinical instability, whereas our task is a binary classification to identify sepsis early.”
Gaussian process regression for forecasting battery state of health (2017, methodology): “They showed that GPs could handle the non-linear data manifested by battery degradation but they concluded that although they were capable of characterizing the uncertainty in the predictions, they lacked long-range predictive capability.”
An Improved Multi-Output Gaussian Process RNN with Real-Time Validation for Early Sepsis Detection (2017, background): “For instance (Ghassemi et al. (2015)) and (Durichen et al. (2015)) use multitask Gaussian processes and (Cheng et al.) use more complex multi-output Gaussian processes, with the focus in all on forecasting future vitals rather than predicting an event.”
Recent Advances in Bayesian Optimization (2022, background): “The need for multi-task learning is ubiquitous across many applications in various fields, such as hyperparameter optimization of machine learning models [235], robotic manipulator inverse dynamics [62], biomedical engineering, and biomedical engineering [62].”
A Multivariate Timeseries Modeling Approach to Severity of Illness Assessment and Forecasting in ICU with Sparse, Heterogeneous Clinical Data (2015, methodology): “Here, we motivate the use of MTGPs and describe the method (source code is available on-line1) that we have adapted for hyperparameter construction (Durichen et al. 2014).”
Remarks on multi-output Gaussian process regression (2018, background): “It has been pointed out that the free-form parameterization allows the elements of Aq to scale freely for each pair of outputs, thus enhancing the ability to well estimate the output correlations [13].”
Sparse Multi-Output Gaussian Processes for Medical Time Series Prediction (2017, methodology): “When modeling medical time series, the SE kernel or the periodic kernel are often used in combination to capture the unknown source-specific smoothness and periodicity of the trajectories of clinical covariates (Stegle et al., 2008; Dürichen et al., 2015).”
Predicting Complications in Critical Care Using Heterogeneous Clinical Data (2016, background): “Use of several recent techniques for modeling sparse, irregularly sampled temporal clinical data (such as [13] and [16]) in combination with multi–view learning might be a useful direction to explore.”
Machine Learning and Decision Support in Critical Care (2016, methodology): “[183], in which multiple time series were fused in a Bayesian nonparametric framework for further improvements in time-series patient monitoring.”