Temporally Multi-Scale Sparse Self-Attention for Physical Activity Data Imputation

TL;DR

This paper introduces a novel sparse self-attention model that leverages domain knowledge to effectively impute missing hourly step count data from wearable sensors, capturing multi-scale temporal patterns.

Contribution

It presents a large-scale dataset and a new domain-informed sparse self-attention approach tailored for multi-scale temporal data imputation.

Findings

The proposed model outperforms baseline methods in imputation accuracy.

Ablation studies confirm the importance of domain knowledge and multi-scale attention.

The dataset enables robust evaluation of imputation techniques.

Abstract

Wearable sensors enable health researchers to continuously collect data pertaining to the physiological state of individuals in real-world settings. However, such data can be subject to extensive missingness due to a complex combination of factors. In this work, we study the problem of imputation of missing step count data, one of the most ubiquitous forms of wearable sensor data. We construct a novel and large scale data set consisting of a training set with over 3 million hourly step count observations and a test set with over 2.5 million hourly step count observations. We propose a domain knowledge-informed sparse self-attention model for this task that captures the temporal multi-scale nature of step-count data. We assess the performance of the model relative to baselines and conduct ablation studies to verify our specific model designs.