Date of Award
Doctor of Philosophy (PhD)
Computational Biology and Bioinformatics
The past decade has witnessed remarkable progress in biomedical informatics and its related fields: the development of high-throughput technologies in genomics, the mass adoption of electronic health records systems, and the AI renaissance largely catalyzed by deep learning. Deep learning has played an undeniably important role in our attempts to reduce the gap between the exponentially growing amount of biomedical data and our ability to make sense of them. In particular, the two main pillars of this dissertation---natural language processing and graph representation learning---have improved our capacity to learn useful representations of language and structured data to an extent previously considered unattainable in such a short time frame. In the context of clinical data, characterized by its notorious heterogeneity and complexity, natural language processing and graph representation learning have begun to enrich our toolkits for making sense and making use of the wealth of biomedical data beyond rule-based systems or traditional regression techniques. This dissertation comes at the cusp of such a paradigm shift, detailing my journey across the fields of biomedical and clinical informatics through the lens of natural language processing and graph representation learning. The takeaway is quite optimistic: despite the many layers of inefficiencies and challenges in the healthcare ecosystem, AI for healthcare is gearing up to transform the world in new and exciting ways.
Chang, David, "Natural Language Processing and Graph Representation Learning for Clinical Data" (2021). Yale Graduate School of Arts and Sciences Dissertations. 26.