Value-based cognition: reward, computation, learning and problem-solving

Many of our daily decisions are related to value. Choosing between food options, doing
mathematic calculations, and learning something new, all of these procedures are
linked to value in different ways. In a series of research, we used neuroimaging (fMRI
and MEG) in humans and single-neuron recordings in rhesus macaque monkeys, to
reveal the neural mechanism of how values are involved in decision-making. We tested
whether there could be neural activities matching economic theories (such as the
expected utility theory) and computer science concepts (such as dynamic
programming). In this talk, I will describe these concept-driven neural signatures in
value-based cognition, namely in multi-attribute choices, risky decision-making, value
learning, and more sophisticated problem-solving.