Annotating sleep states in children from wrist-worn accelerometer data using Machine Learning
This work addresses the challenge of precise and scalable sleep annotation for researchers studying children's sleep patterns, representing an incremental application of existing methods to a specific domain.
The paper tackled the problem of automating sleep state annotation in children from wrist-worn accelerometer data by modeling it with various machine learning techniques, including support vectors, boosting, ensemble methods, LSTMs, and Region-based CNNs, and evaluated them using the Event Detection Average Precision (EDAP) score to compare predictive power and performance.
Sleep detection and annotation are crucial for researchers to understand sleep patterns, especially in children. With modern wrist-worn watches comprising built-in accelerometers, sleep logs can be collected. However, the annotation of these logs into distinct sleep events: onset and wakeup, proves to be challenging. These annotations must be automated, precise, and scalable. We propose to model the accelerometer data using different machine learning (ML) techniques such as support vectors, boosting, ensemble methods, and more complex approaches involving LSTMs and Region-based CNNs. Later, we aim to evaluate these approaches using the Event Detection Average Precision (EDAP) score (similar to the IOU metric) to eventually compare the predictive power and model performance.