Machine Learning and Visualizations for Time-Series Healthcare Data
Sayantani Basu · 2023
Time-series healthcare data is difficult to analyze and visualize due to: (1) variations in data over time, (2) different modes of physiological signals (genetic data, behavioral data, etc.), and (3) difficulty in analyzing high-dimensional samples in cases where the data is diverse in nature. My research focuses on analyzing such data using machine learning and visualizations. The research questions I wish to address through my dissertation are: (1) How can multidimensional healthcare data be analyzed using machine learning to identify key moments of significance? and (2) What are the methods of visualizing these key moments in an interpretable manner to aid healthcare professionals and caregivers?