I am a quantitative psychologist focusing on developing novel methods to study complex psychological processes and how these psychological processes impact important outcomes, such as mental health. The methods I have been developing focus on dynamical system methods, control theory, network analysis, and machine learning. My Ph.D. training is Quantitative Methods from the Human Development and Family Studies Department at Penn State, and I am currently a postdoctoral researcher at Stanford University (Psychology Department).
My research work capitalizes on intensive longitudinal data that records how each person responds to emotional and social experiences as he/she goes about their lives. I model this complex process as an intertwined dynamic network of interactive components (emotion, behavior, physiological), and detect when a specific network structure poses a risk for the person's mental health. My previous work has found that network structures like positive feedback loops between negative emotions and social behaviors can extend the time to recover from elevated sad mood, and subsequently increase the level of depressive symptoms.
Once I can identify such risks in complex dynamic processes, I begin to design control to direct the person out of the dangerous scenarios. Inspired by systems biology literature, I use the Boolean network method to inform personalized intervention for emotion regulation dynamics. In the future, I am interested in exploring possibilities to integrate network control with digital devices and deliver prompt alerts to reduce risks of mental illness.
Digital devices also allow for passive data collection with much denser measurement and higher precision. Artificial intelligence (AI), in conjunction with these data, enables modeling of psychological processes, including emotional states, cognitive load, physical activity, and social behaviors. I am excited to explore the possibilities of applying AI methods, such as reinforcement learning, to research how people adapt to their environment.