Ma researches how the human brain represents and processes
uncertainty. A large portion of his academic work is devoted to the construction of
Bayesian inference models which describe how an observer arrives at beliefs about things in the world from noisy information. This modeling spans the range of describing the behavior of an observer, for example how an observer might infer two sensory inputs arise from a common source, to the activity of a population of neurons implementing the Bayesian operations. A complementary line of inquiry is studying encoding strategies in
working memory, specifically highlighting the relationships between the role of noisy representations of objects in the brain and the number that can be recalled correctly. Broadly, his modeling focuses can be described as examining encoding models, decision rules, and probabilistic computations. In recent years, his work has focused on planning and thinking ahead in complex decision problems. == Non-academic activities ==