Big data approaches to understanding the human brain
Jack L. Gallant, Ph.D.
Professor, Department of Psychology
University of California-Berkeley
The human brain is the most sophisticated computer processing system known, capable of impressive feats of recognition, evaluation and prediction under challenging natural conditions. Reverse-engineering the human brain might enable us to design artificial systems with the same capabilities. Efforts to reverse-engineer the brain are inevitably limited by the technology available for recording brain activity and by the amount of data that can be recorded. Recent advances and ongoing development of human neuroimaging offer an opportunity to use a data-driven, big-data approach to the reverse engineering problem. In this talk I will summarize ongoing work in my laboratory that uses a data-driven system identification approach to tackle this reverse-engineering problem. Our approach consists of several broad stages.
As neuroimaging technology advances and more data can be recorded from the brain, the quality of the fit models and the quality of brain decoding will increase dramatically, and data-driven approaches to brain mapping will become increasingly powerful and important.