Temporal-Clustering Invariance in Irregular Healthcare Time Series

Abstract

Electronic records contain sequences of events, some of which take place all at once in a single visit, and others that are dispersed over multiple visits, each with a different timestamp. We postulate that fine temporal detail, e.g., whether a series of blood tests are completed at once or in rapid succession should not alter predictions based on this data. Motivated by this intuition, we propose models for analyzing sequences of multivariate clinical time series data that are invariant to this temporal clustering. We propose an efficient data augmentation technique that exploits the postulated temporal-clustering invariance to regularize deep neural networks optimized for several clinical prediction tasks. We introduce two techniques to temporally coarsen (downsample) irregular time series: (i) grouping the data points based on regularly-spaced timestamps; and (ii) clustering them, yielding irregularly-paced timestamps. Moreover, we propose a MultiResolution Ensemble (MRE) model, improving predictive accuracy by ensembling predictions based on inputs sequences transformed by different coarsening operators. Our experiments show that MRE improves the mAP on the benchmark mortality prediction task from 51.53% to 53.92%.

References in corpus (11)

• Adam: A Method for Stochastic Optimization
• Explaining and Harnessing Adversarial Examples
• On the Convergence of Adam and Beyond
• Group Equivariant Convolutional Networks
• Semi-supervised Sequence Learning
• On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups
• Modeling Missing Data in Clinical Time Series with RNNs
• An Improved Multi-Output Gaussian Process RNN with Real-Time Validation for Early Sepsis Detection
• Temporal Convolutional Neural Networks for Diagnosis from Lab Tests
• Interpolation-Prediction Networks for Irregularly Sampled Time Series
• Learning from Clinical Judgments: Semi-Markov-Modulated Marked Hawkes Processes for Risk Prognosis

Cited by in corpus (10)

• SCINet: Time Series Modeling and Forecasting with Sample Convolution and Interaction
• A Review of Deep Learning Methods for Irregularly Sampled Medical Time Series Data
• Set Functions for Time Series
• Explaining Deep Classification of Time-Series Data with Learned Prototypes
• A Survey on Principles, Models and Methods for Learning from Irregularly Sampled Time Series
• Evaluating Progress on Machine Learning for Longitudinal Electronic Healthcare Data
• Real-time Prediction for Mechanical Ventilation in COVID-19 Patients using A Multi-task Gaussian Process Multi-objective Self-attention Network
• Improving Irregularly Sampled Time Series Learning with Dense Descriptors of Time
• Discovering Invariances in Healthcare Neural Networks
• Uncertainty-Gated Stochastic Sequential Model for EHR Mortality Prediction