A seminar will be given by Research Assistant Atakan Aygün on Physics-Informed Neural Networks (PINNs), an emerging methodology for addressing forward and inverse problems in partial differential equations.
The talk will introduce the mathematical foundations of PINNs, discuss how physical laws are embedded into neural network architectures, and highlight their advantages over traditional numerical approaches. Applications across engineering domains will also be presented, emphasizing the growing relevance of PINNs in mechanical engineering research.
Date: November 21st, 14:30 Place: E200, Department of Mechanical Engineering, METU
