Please note: This PhD seminar will take place online.
Ehsan Ganjidoost, PhD candidate
David R. Cheriton School of Computer Science
Supervisor: Professor Jeff Orchard
The seminar will provide a comprehensive overview of key concepts in perception and learning modelling, focusing on the predictive coding and free-energy framework as an extension. The free-energy framework offers a powerful approach to understanding how organisms perceive and learn from their environment. It integrates principles from Bayesian inference and predictive coding, allowing for the modelling of complex cognitive processes.
The seminar will cover the theoretical foundations of the free-energy framework, its applications in various domains, and the computational mechanisms involved. Additionally, the relationship between predictive coding and backpropagation, a fundamental algorithm in neural network training, will be explored, shedding light on the connection between biological and artificial learning systems.