RA Position Available in Deep Learning (particularly, Online or Self-Supervised Learning) for Wearable Sensor Data
We are looking for a new PhD student to join our research group in Fall 2023. Please read the project description and necessary skillsets below. If interested, please reach out to Prof. Lee at silee@cs.umass.edu along with your CV. You will also need to apply to the Ph.D. program in the College of Information and Computer Sciences (Deadline: 12/15). Applicants are encouraged to specify that they are interested in working with Prof. Lee in their SOP. Please also visit the lab webpage: http://www.ahhalab.org/
Description
Explicitly focusing on individuals undergoing long-term rehabilitation – such as stroke survivors and TBI survivors – we are interested in improving our understanding of how individuals perform functional movements outside the clinic using various on-body and ambient sensing technologies. We are looking for students who are passionate about designing novel machine-learning algorithms to extract clinically relevant information from noisy data obtained from patients’ naturalistic environments.
Related published papers from this project include the following:
- Estimating Upper-Limb Impairment Level in Stroke Survivors using Wearable Inertial Sensors and a Minimally-Burdensome Motor Task
Brandon Oubre, Jean-Francois Daneault, Hee-Tae Jung, Kallie Whritenour, Jose Garcia Vivas Miranda, Joonwoo Park, Taejyeong Ryu, Yangsoo Kim, Sunghoon Ivan Lee, IEEE Transactions on Neural Systems & Rehabilitation Engineering (IEEE TNSRE), March 2020, [PDF] - Analysis of Gait Sub-Movements to Estimate Ataxia Severity using Ankle Inertial Data
Juhyeon Lee, Brandon Oubre, Jean-Francois Daneault, Christopher D. Stephen, Jeremy D. Schmahmann, Anoopum S. Gupta*, and Sunghoon Ivan Lee*, IEEE Transactions on Biomedical Engineering (IEEE TBME), July 2022, [PDF].
Knowledge, Skills, and Abilities
The Ph.D. students will closely interact with Prof. Lee and other team members on a weekly basis. They are expected to have excellent oral and written communication, creativity for algorithm and/or qualitative analysis, and teamwork spirit. We prefer candidates with the following skills:
- Strong understanding and prior experience in conventional feature-based supervised learning. I prefer prior experience with time-series data analysis, such as accelerometer or other physiological data, but it is not mandatory.
- Strong understanding of the fundamentals of machine learning and its application to domain-specific data. Having experience in analyzing wearable time-series data, such as accelerometer or other physiological data, to construct regression or classification models is a plus.
- Strong understanding or prior experience in deep learning-based machine learning models, particularly online & self-supervised learning algorithms. Its application to time-series analysis is a plus.
Note that we do not require applicants to have domain-specific, clinical knowledge in rehabilitation medicine. I welcome students with diverse backgrounds who have strong interests in the subject matter & enthusiasm to advance healthcare using technologies.