CAREER: Time-resolved decoding of explore/exploit computations in the human brain

University Of New Mexico, Albuquerque NM

Investigators

Abstract

The exploration-exploitation tradeoff is a dilemma we frequently face in choosing between options. This tradeoff happens all the time in everyday life - do we ‘exploit’ what we already know, or ‘explore’ something new and unfamiliar? Do we choose to go to our favorite restaurant (exploitation) or try out a new restaurant (exploration)? Other examples include deciding where to go for vacation, which friends to talk to at a party, or what foods to order at the restaurant. While exploring new places, new friends, new foods is key to figuring out what we like and can lead to the formation of new preferences, it comes at the cost of foregoing the immediate value of familiar options. This project seeks to understand how the decision to explore or exploit takes shape in the human brain. In order to understand the neural basis of decision making, two non-invasive, complementary neuroimaging techniques can be used to analyze brain activity while people make the decision to explore or exploit. One approach (fMRI or functional magnetic resonance imaging) has superb spatial resolution, and can tell where brain activity is localized during decision-making. Another approach (MEG or magnoencephalography) has excellent temporal resolution and can pinpoint precisely when brain signals occur while people make explore-exploit decisions. Relatively little is known about the timing of these decisions in the brain. Utilizing these neuroimaging techniques, experiments measure the spatiotemporal patterns of brain activity in situations where people are trying to gain as many profits as possible, as well as in contrasting situations where people are trying to avoid losses. This will allow an experimental test of the hypothesis that the brain circuits involved when people are trying to maximize profits through explore-exploit decisions are distinct from the circuits involved when they are trying to minimize losses. Using computational modeling, the project will also study the accuracy of decoding exploration and exploitation decisions from MEG brain data under different conditions and at different points in a decision trial. Additional research will use a technique (TMS or transcranial magnetic stimulation) for safely stimulating the brain to provide a ‘cause-and-effect’ test of predictions about how the brain controls explore-exploit decisions. How the human brain chooses between familiarity and novelty during decision-making is a long-standing mystery in human brain and decision sciences, and this research project is likely to greatly increase our knowledge of the neural mechanisms underlying decision making. These studies will also have significant societal implications in educational and clinical settings and in a broad array of contexts where insights from neuroscience may encourage and help people to make better choices. The scientific research in this project will be accompanied by an outreach program to teach computer programming and STEM skills to high school students, and introduce them to the excitement of neuroscience research. This project is jointly funded by the Cognitive Neuroscience Program in SBE, and the Established Program to Stimulate Competitive Research (EPSCoR). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.