Dr. Deborah Duong is Director or AI Development at Rejuve and Director of Network Analytics at Singularity Net. The focus of her research is Complex Adaptive Systems, on the boundary between AI and Computational Social Science. She wrote the world's first Intelligent Agent Based Social Simulation for her University of Alabama, Birmingham Masters in Computer Science 1991, "A System of IAC Networks as the Basis for Self Organization in a Sociological Dynamical System Simulation" (published in Behavioral Science) and obtained her doctorate in Computational Sciences and Informatics with a major in Computational Social Science from GMU in 2004. Her dissertation, published in JASSS (but based on her 1996 NIST paper),"SISTER: A Symbolic Interactionist Simulation of Trade and Emergent Roles", demonstrates micro macro integration of social phenomena from multiple reinforcement learning agents. In her models, from agents with individual neural networks or genetic algorithms for minds that interact according to micro sociology emerged the macro patterns of status symbols, racism, social class, a division of labor, price,and the emergence of money from barter.
SISTER formed the basis of a decade of work at the US Office of the Secretary of Defense and US Army, where she applied her Nexus neural and genetic agents to model corruption and popular support for the first Irregular Warfare Analytical Baseline, the study used to allocate the budget to the services, as well as for Irregular Warfare Tactical Wargame, a study that won the highest US Army analysis award. She later turned her attention to health policy simulation in private industry, modelling diseases responses to treatments with deep neural nets and the effects of Affordable Care Act's bundled payment program on incentives for treatments with evolutionary computation reinforcement learning agents. At Singularity Net, she applied her SISTER CMA-ES social agents to AutoML, and Bayesian agents to a reputation system simulation of an Amazon marketplace.