No Imputation Needed: A Switch Approach to Irregularly Sampled Time Series

TL;DR

This paper introduces SLAN, a novel LSTM-based model that handles irregularly sampled time series without imputation, effectively capturing local and global summaries by dynamically adapting to sensor data.

Contribution

SLAN is the first model to model ISTS without imputation, using switches to adapt architecture based on sensor measurements, avoiding bias from missing data assumptions.

Findings

SLAN outperforms existing methods on MIMIC-III and Physionet 2012 datasets.

It effectively captures local sensor summaries and maintains a global state.

The approach eliminates the need for imputation, reducing bias and improving performance.

Abstract

Modeling irregularly-sampled time series (ISTS) is challenging because of missing values. Most existing methods focus on handling ISTS by converting irregularly sampled data into regularly sampled data via imputation. These models assume an underlying missing mechanism, which may lead to unwanted bias and sub-optimal performance. We present SLAN (Switch LSTM Aggregate Network), which utilizes a group of LSTMs to model ISTS without imputation, eliminating the assumption of any underlying process. It dynamically adapts its architecture on the fly based on the measured sensors using switches. SLAN exploits the irregularity information to explicitly capture each sensor's local summary and maintains a global summary state throughout the observational period. We demonstrate the efficacy of SLAN on two public datasets, namely, MIMIC-III, and Physionet 2012.