My research leverages techniques from AI, machine learning, and computational cognitive science to predict human behavior under risk and uncertainty. Explore this site to learn about my experience in academia and industry. Thanks for visiting.
Please feel free to contact me with questions about my research or if you're interested in collaborating.DOWNLOAD CV
I am interested in improving decision-making through decision support systems.
I have experience developing computational games that quantify human performance.
I enjoy developing machine learning and AI algorithms across several domains.
I have a keen interest in modeling human performance under uncertainty and risk.
At MIT, Dr. Schlicht presented his research to the Computational Cognitive Science Group, Drazen Prelec's Neuroeconomics Group, and the Perceptual Sciences Group.
At Harvard, Dr. Schlicht gave presentations at the Medical School and to the Vision Sciences Group.
Dr. Schlicht presentated his research to both the Shimojo and Andersen Laboratories.
At Stanford, Dr. Schlicht presented his work to the Intelligent Systems Laboratory.
Dr. Schlicht gave an invited talk at the SAMSI Summer Program on Transportation Statistics.
This page contains a summary of my experience across industry and academia.
Dr. Schlicht is currently working as the Founder of the BrainCallus Gaming Project and is responsible for all technical and business aspects of the company. The BrainCallus Gaming Project is an effort seeking to improve psychiatric decision-making by leveraging a combination of computational gaming, machine learning and cognitive science. He is currently developing the Brain Barn Series of games that contain modules capable of quantifying player perception, decision-making and movements.
Dr. Schlicht was the Founder of the Computational Cognition Group (C2-g), LLC and was responsible for all technical and business aspects of the company. During his tenure as founder, he gained attention for leveraging multifidelity methods and computational gaming to design decision-support systems. He also developed a novel model to improve the prediction of NFL outcomes by exploiting oddsmaker decision biases.
Dr. Schlicht returned to the University of Minnesota to conduct research in the HumanFIRST Lab where he developed machine learning algorithms (e.g., Bayesian Networks, Support Vector Machine regression, Binary classification with Lasso) to predict human driving behavior. These models were then used to estimate the risk associated with candidate transportation technology by using the predictive models in multifidelity simulation.
At Medtronic, Dr. Schlicht was part of a team that was responsible for developing next-generation Deep-Brain-Stimulation devices to help treat diseases, such movement disorders.
Dr. Schlicht was a researcher at MIT Lincoln Laboratory conducting research related to national security. He was responsible for developing a novel model to predict the decisions of interacting humans. The model defined a quantitative method to combine the results from low-fidelity simulations (e.g., novice in an online simulator) with high-fidelity simulations (e.g., expert in an immersive simulator) to evaluate when inexpensive low-fidelity data can be used to as a proxy for expensive high-fidelity simulations. Moreover, he was part of an effort to use Serious Games as a means to develop quantitative models of operational decision-making.
Dr. Schlicht was a Cognitive Scientist at Aptima and led several SBIR and STTR efforts on projects related to national security. In his brief time at Aptima, he was awarded one OSD contract for a biologically-inspired approach to automated scene estimation (BIS-E), in addition to successfully securing one patent for quantifying human reactions to communications.
While a postdoctoral researcher between Harvard University and Caltech, Dr. Schlicht developed a low-fidelity game to quantitatively investigate human decision-making in a competitive (zero-sum) task. This research received an enormous amount of public interest and has been covered by several major media outlets (see list below), and resulted in a publication that was rated in the Top 5 Percent of all research output according to metrics by Altmetric.
Dr. Schlicht has instructed several undergraduate courses at the University of Minnesota, Wellesley College, and Harvard University. In 2009, he was awarded the Certificate of Teaching Distinction from Harvard University.
Major: Cognitive and Brain Sciences; Minor: Human Factors
Major: Psychology; Minor: Biology
This page contains an assortment of publications and preprints of my work. See my CV for a full list.
Schlicht, E.J. (2017). Exploiting oddsmaker bias to improve the prediction of NFL outcomes. arXiv: Statistical Applications.
Schlicht, E.J. & Morris, N. (2017). Estimating the risk associated with candidate transportation technology through multifidelity simulation. arXiv: Statistical Applications.
Schlicht, E.J., Lee, R., Wolpert, D., Kochenderfer, M. , & Tracey, B. (2012). Predicting the behavior of interacting humans by fusing data from multiple sources. In the Proceedings of the Twenty-Eighth Conference of Uncertainty in Artificial Intelligence, (UAI-2012). [30% Acceptance Rate]
Schlicht, E.J., Shimojo, S., Camerer, C., Battaglia, P.R., & Nakayama, K. (2010). Human wagering behavior depends on opponents faces, PLoS ONE, 5(7): e11663. doi:10.1371/journal.pone.0011663.
Schlicht, E.J., & Schrater, P.R. (2007). Impact of coordinate transformation uncertainty on human sensorimotor control. Journal of Neurophysiology, 97(6), pp. 4203-14.