On January 10th, Professor Haibo He, the Chief Editor of IEEE Transactions on Neural Network and Learning System from University of Rhode Island, and Professor Yaochu Jin, the Chief Editor of IEEE Transactions on Cognitive and Developmental Systems from University of Surrey, accepted the invitations from the School of Information Science and Engineering of ECUST, and visited ECUST. They delivered two academic lectures entitled “Machine Learning: From Shallow to Deep” and “Some Recent Advances in Federated Learning”, respectively. The lectures were hosted by Professor Du Wenli, the Dean of the School of Information Science and Engineering, which are very welcomed by the faculty members and students.

Professor Haibo He started by talking about the amazing achievements realized by artificial intelligence in various fields and then focused on deep learning which is the core of artificial intelligence. After that, he elaborately explained why we need deep learning, how to make a learning model deep, and how to apply deep learning models in real applications. In addition, he introduced their latest research results on the applications of Smart Grid and Robotics.

Professor Yaochu Jin started by talking about the background and several major challenges  in Federated Learning, such as the data privacy and communication limitation. He also shared their most recent researches on addressing the above challenges by passing only part of the parameters to the server, transferring knowledge from history, and reducing the model complexity using a multi-objective evolutionary algorithm. Professor Yaochu Jin answered the questions from audiences in detail during the lively question and answer session.