Cortical Sensing – a novel approach to studying the neural mechanisms underlying learning.
Learning is the process by which outer experience induce change on an organism behavior. As the ability to learn underlies a vast range of behaviors, a central goal of modern day neuroscience is to understand the neural basis of learning, a goal which we are still far from achieving. Learning is intuitively viewed as the brain adapting to external stimuli. However, since all such stimuli are translated into neural activity, learning is in fact the brain adapting to its own activity.
A closely related discipline of study is machine learning, namely the field studying the mathematical and algorithmic principles of learning. From a theoretical perspective, machine learning provides a complete mathematical analysis of learning processes and can also be used to make concrete predictions, and to analyze observed processes to understand the underlying mechanisms. Therefore, given the above, one can view the brain adaption to its own activity as “solving” machine learning problems of recognizing neural activity.
In my research, I will focus on understanding which patterns of neural activity can be learned – “sensed” by the brain – and to find the neuronal mechanisms underlying such activity. In short – activating cortical patterns in a freely moving rodent and studying its ability to learn different patterns. To address this problem, I’m constructing a setup where an animal is taught to learn specific patterns of cortical activity. I plan on performing a wide range of experiments to test learning of different activity patterns. The design and analysis of these experiments will be based on machine learning theory.
Dan Maoz, Abraham Loeb, Yossi Shvartzvald, Monika Sitek, Michael Engel, Flavien Kiefer, Marcin Kiraga, Amir Levi, Tsevi Mazeh, Michal Pawlak, R. Michael Rich, Lev Tal-Or, Lukasz Wyrzykowski: Fast radio bursts: the observational case for a Galactic origin. MNRAS, 2016
Research Categories: Behavioral Neuroscience, Electrophysiology, Machine learning, Optogenetics