Research

My research primarily focuses on performing machine learning on incomplete data, especially in the case where the available data is biased. This includes projects on learning from data with partial or missing labels, as well as debiasing datasets that underrepresent certain demographics. My main focus is on developing rigorous models of the annotation process itself for correcting and debiasing missing labels (AAAI 2022, SDM 2022), and applying generative modeling to correct for biased data samples (Big Data 2022).

My work is driven by the development of Human Context Recognition (HCR) systems that identify the context (i.e., physical activities and state) of individuals using mobile sensor data. I focus primarily on developing HCR systems that are beneficial to downstream mobile healthcare applications (IEEE Pervasive Computing 2021).