CARRNN: A Continuous Autoregressive Recurrent Neural Network for Deep Representation Learning from Sporadic Temporal Data

TL;DR

CARRNN is a novel deep learning model that effectively captures sporadic and asynchronous temporal patterns in multivariate data, improving prediction accuracy in healthcare applications like Alzheimer's progression and ICU mortality.

Contribution

The paper introduces CARRNN, a continuous-time autoregressive RNN that models irregular temporal data end-to-end, outperforming existing RNN-based methods in healthcare time-series prediction.

Findings

CARRNN achieves the lowest prediction errors in Alzheimer's progression modeling.

CARRNN outperforms state-of-the-art GRU and LSTM models in ICU mortality prediction.

The model effectively handles irregular and asynchronous data in multivariate time-series tasks.

Abstract

Learning temporal patterns from multivariate longitudinal data is challenging especially in cases when data is sporadic, as often seen in, e.g., healthcare applications where the data can suffer from irregularity and asynchronicity as the time between consecutive data points can vary across features and samples, hindering the application of existing deep learning models that are constructed for complete, evenly spaced data with fixed sequence lengths. In this paper, a novel deep learning-based model is developed for modeling multiple temporal features in sporadic data using an integrated deep learning architecture based on a recurrent neural network (RNN) unit and a continuous-time autoregressive (CAR) model. The proposed model, called CARRNN, uses a generalized discrete-time autoregressive model that is trainable end-to-end using neural networks modulated by time lags to describe the changes caused by the irregularity and asynchronicity. It is applied to multivariate time-series regression tasks using data provided for Alzheimer's disease progression modeling and intensive care unit (ICU) mortality rate prediction, where the proposed model based on a gated recurrent unit (GRU) achieves the lowest prediction errors among the proposed RNN-based models and state-of-the-art methods using GRUs and long short-term memory (LSTM) networks in their architecture.