Chathika is a Research Scientist in the Learning Systems Group at Oak Ridge National Laboratory. His research lies at the conjunction of Complex Adaptive Systems, Evolutionary Computation, and Machine Learning. Specifically, his work focuses on understanding complex social systems through explainable AI. Some of his current projects include modeling and analysis of the adverse impacts that COVID-19 disinformation have on health outcomes and weakly supervised learning models for information extraction.
Prior to ORNL, Chathika worked as a Postdoctoral Associate under the supervision of Dr. Una-May O'Reilly at the Any-Scale Learning for All group at MIT-CSAIL. Here, he led the construction of simulation-based force-on-force wargames for Red Force brigade planning with novel tactics, techniques and procedures for the DARPA COMBAT program. He also contributed to the development of agent-based models for operational assessment of safe return-to-work policies as part of MIT's Quest for Intelligence program.
Chathika's PhD work was conducted under the supervision of Dr. Ivan Garibay at the Complex Adaptive Systems Laboratory at the University of Central Florida, and supported by the DARPA SocialSim program. His dissertation work centered around the development of an explainable AI framework, Evolutionary Model Discovery, for inferring importance of individual-scale factors of decision-making from complex social phenomena. Chathika also led the development of a model for at-scale simulation of cross-platform information spread through social media as part of the SocialSim program.
Chathika's experience in industry includes appointments at NBC Universal Studios, where he developed agent-based models of pedestrian dynamics, and at SimCentric Technologies, where he developed crowd simulation for military training simulations.