Filters








812 Hits in 8.4 sec

Bayesian Recurrent Framework for Missing Data Imputation and Prediction with Clinical Time Series [article]

Yang Guo, Zhengyuan Liu, Pavitra Krishnswamy, Savitha Ramasamy
2020 arXiv   pre-print
Recent works propose recurrent neural network based approaches for missing data imputation and prediction with time series data.  ...  In this work, we introduce a unified Bayesian recurrent framework for simultaneous imputation and prediction on time series data sets.  ...  Discussion and Future Work We have developed a Bayesian recurrent framework to enable missing data imputation and prediction on clinical time series data sets.  ... 
arXiv:1911.07572v2 fatcat:hqc5rh5nnbeu5eqnr3i2pdbuqy

Applying an Instance-Specific Model to Longitudinal Clinical Data for Prediction

Emily Watt, James W. Sayre, Alex A.T. Bui
2011 2011 IEEE First International Conference on Healthcare Informatics, Imaging and Systems Biology  
., conditional probability tables) are learned in the context of missing data for: 1) a dynamic belief network (DBN), which is used as a population-wide temporal model; and 2) an instant-specific patient  ...  Section IV presents the evaluation results and comparison of the two models' predictive power. Finally, Sections V and VI provide a discussion of the results and conclude with potential future work.  ...  This manuscript was prepared using an OAI public use data set and does not necessarily reflect the opinions or view of the OAI investigators, the NIH, or the private funding partners.  ... 
doi:10.1109/hisb.2011.12 pmid:27570832 pmcid:PMC5001560 dblp:conf/hisb/WattSB11 fatcat:mwbcrd7skzeihpr7hac3ubmwcu

Uncertainty-Aware Variational-Recurrent Imputation Network for Clinical Time Series [article]

Ahmad Wisnu Mulyadi, Eunji Jun, Heung-Il Suk
2020 arXiv   pre-print
In this work, we propose a novel variational-recurrent imputation network, which unifies an imputation and a prediction network by taking into account the correlated features, temporal dynamics, as well  ...  Since the missing value estimates involve the risk of being inaccurate, it is appropriate for the method to handle the less certain information differently than the reliable data.  ...  CONCLUSION In this study, we proposed a novel unified framework consisting of imputation and prediction networks for sparse high-dimensional multivariate time series.  ... 
arXiv:2003.00662v2 fatcat:gkm463rpangm5l7fx7itmaocye

Compound Density Networks for Risk Prediction using Electronic Health Records [article]

Yuxi Liu, Zhenhao Zhang, Shaowen Qin
2022 arXiv   pre-print
Imputation of missing values has been considered an effective approach to deal with this challenge.  ...  The architecture of CDNet enables GRU and MDN to iteratively leverage the output of each other to impute missing values, leading to a more accurate and robust prediction.  ...  We use clinical times series data (e.g., heart rate, glucose) as input [11] .  ... 
arXiv:2208.01320v2 fatcat:pqsw3wqlw5cahjrnu42mkqpari

Quality control, data cleaning, imputation [article]

Dawei Liu, Hanne I. Oberman, Johanna Muñoz, Jeroen Hoogland, Thomas P.A. Debray
2021 arXiv   pre-print
Finally, we introduce alternative methods to address incomplete data without the need for imputation.  ...  We motivate why the imputation of RWD may require additional efforts to avoid bias, and highlight recent advances that account for informative missingness and repeated observations.  ...  Acknowledgements This project has received funding from the European Union's Horizon 2020 research and innovation programme under ReCoDID grant agreement No 825746.  ... 
arXiv:2110.15877v1 fatcat:tzswjt4kerd7zegbgt5mdm23gi

Remiod: Reference-based Controlled Multiple Imputation of Longitudinal Binary and Ordinal Outcomes with non-ignorable missingness [article]

Tony Wang, Ying Liu
2022 arXiv   pre-print
Missing data on response variables are common in clinical studies. Corresponding to the uncertainty of missing mechanism, theoretical frameworks on controlled imputation have been developed.  ...  In this paper, we introduce the R package remiod, which utilizes the Bayesian framework to perform imputation in regression models on binary and ordinal outcomes.  ...  Based on an assumption of missing mechanism, replacing missing data with substituted values, i.e. imputation, becomes popular with theoretical advancement in methodology for analyzing data with missingness  ... 
arXiv:2203.02771v1 fatcat:32jvcggysbfidivqknoddkhbyy

Deep Sensing: Active Sensing using Multi-directional Recurrent Neural Networks

Jinsung Yoon, William R. Zame, Mihaela van der Schaar
2018 International Conference on Learning Representations  
For every prediction we might wish to make, we must decide what to observe (what source of information) and when to observe it.  ...  To infer missing measurements, we develop a Multi-directional Recurrent Neural Network (M-RNN).  ...  ACKNOWLEDGMENTS This work was supported by the Office of Naval Research (ONR) and the NSF (Grant number: ECCS1462245, ECCS1533983, and ECCS1407712).  ... 
dblp:conf/iclr/YoonZS18 fatcat:pchgux25l5djneeya4c75s3taq

Mixture-based Multiple Imputation Model for Clinical Data with a Temporal Dimension [article]

Ye Xue, Diego Klabjan, Yuan Luo
2019 arXiv   pre-print
The problem of missing values in multivariable time series is a key challenge in many applications such as clinical data mining.  ...  Although many imputation methods show their effectiveness in many applications, few of them are designed to accommodate clinical multivariable time series.  ...  Imputation Framework In many predictive tasks on temporal clinical data, time series are often aligned into the same-length sequences to derive more robust patient phenotypes through matrix decomposition  ... 
arXiv:1908.04209v2 fatcat:tyw4e5naczfudkzytquj4prfpa

Clairvoyance: A Pipeline Toolkit for Medical Time Series

Daniel Jarrett, Jinsung Yoon, Ioana Bica, Zhaozhi Qian, Ari Ercole, Mihaela van der Schaar
2021 International Conference on Learning Representations  
At the same time, medical time-series problems in the wild are challenging due to their highly composite nature: They entail design choices and interactions among components that preprocess data, impute  ...  Time-series learning is the bread and butter of data-driven clinical decision support, and the recent explosion in ML research has demonstrated great potential in various healthcare settings.  ...  ACKNOWLEDGMENTS We would like to thank the reviewers for their generous and invaluable comments and suggestions.  ... 
dblp:conf/iclr/JarrettYBQES21 fatcat:ltp7ijb24rhqnccpby74cooflq

Editorial: Predictive Intelligence in Biomedical and Health Informatics

E. Adeli, S. H. Rekik, S. H. Park, D. Shen
2020 IEEE journal of biomedical and health informatics  
); predicting lesion evolution; predicting missing data (e.g., data imputation or data completion problems).  ...  They proposed a multi-task recurrent neural network with attention mechanisms to predict patients' hospital mortality, using reconstruction of patient physiological time series as an auxiliary task.  ... 
doi:10.1109/jbhi.2019.2962852 fatcat:mp2kymg7yjb7ziukia7kjgjula

Probabilistic Machine Learning for Healthcare [article]

Irene Y. Chen, Shalmali Joshi, Marzyeh Ghassemi, Rajesh Ranganath
2020 arXiv   pre-print
We consider challenges in the predictive model building pipeline where probabilistic models can be beneficial including calibration and missing data.  ...  Beyond predictive models, we also investigate the utility of probabilistic machine learning models in phenotyping, in generative models for clinical use cases, and in reinforcement learning.  ...  Krishnan, Peter Schulam, and Pete Szolovits for helpful and useful feedback. This work was supported in part by a CIFAR AI Chair at the Vector Institute (MG) and Microsoft Research (MG).  ... 
arXiv:2009.11087v1 fatcat:htosfeqvhndvfmlmud2pvl3nsy

Machine Learning and Decision Support in Critical Care

Alistair E. W. Johnson, Mohammad M. Ghassemi, Shamim Nemati, Katherine E. Niehaus, David Clifton, Gari D. Clifford
2016 Proceedings of the IEEE  
This paper discusses the issues of compartmentalization, corruption, and complexity involved in collection and preprocessing of critical care data.  ...  [183] , in which multiple time series were fused in a Bayesian nonparametric framework for further improvements in time-series patient monitoring.  ...  Multiple imputation, a technique which involves repeatedly imputing plausible values for missing data and averaging over many instances of imputation [108] , [109] , has received wide praise among the  ... 
doi:10.1109/jproc.2015.2501978 pmid:27765959 pmcid:PMC5066876 fatcat:7i6wi65qgjbapjjznk2nioz32y

Sensitivity analysis for clinical trials with missing continuous outcome data using controlled multiple imputation: A practical guide

Suzie Cro, Tim P. Morris, Michael G. Kenward, James R. Carpenter
2020 Statistics in Medicine  
This tutorial provides an overview of controlled multiple imputation (MI) techniques and a practical guide to their use for sensitivity analysis of trials with missing continuous outcome data.  ...  Missing data due to loss to follow-up or intercurrent events are unintended, but unfortunately inevitable in clinical trials.  ...  Andrew Vickers and the acupuncture trial team for the use of their data within this tutorial. We would also like to thank Dr. Sara Schroter and the reviewer trial team for the their data.  ... 
doi:10.1002/sim.8569 pmid:32419182 fatcat:kshohmakhbhrvoqusbrcva6yoy

Gradient Importance Learning for Incomplete Observations [article]

Qitong Gao, Dong Wang, Joshua D. Amason, Siyang Yuan, Chenyang Tao, Ricardo Henao, Majda Hadziahmetovic, Lawrence Carin, Miroslav Pajic
2022 arXiv   pre-print
We test the approach on real-world time-series (i.e., MIMIC-III), tabular data obtained from an eye clinic, and a standard dataset (i.e., MNIST), where our imputation-free predictions outperform the traditional  ...  missing values without imputation.  ...  Since the focus of this work is principally on time-series data, recurrent neural networks will receive significant attention.  ... 
arXiv:2107.01983v4 fatcat:qumjhvlnlvh4bbeztrijfurlqi

Early Recognition of Sepsis with Gaussian Process Temporal Convolutional Networks and Dynamic Time Warping [article]

Michael Moor and Max Horn and Bastian Rieck and Damian Roqueiro and Karsten Borgwardt
2020 arXiv   pre-print
Meanwhile, despite decades of clinical research, robust biomarkers for sepsis are missing.  ...  Our deep learning model employs a temporal convolutional network that is embedded in a Multi-task Gaussian Process Adapter framework, making it directly applicable to irregularly-spaced time series data  ...  For example, to evaluate the prediction horizon at 3 h in advance, for each encounter, the model (and imputation scheme) is only provided with input data up until that moment.  ... 
arXiv:1902.01659v4 fatcat:24b4iwpih5d5rijci4chl5ybie
« Previous Showing results 1 — 15 out of 812 results