Andreas Moschovos is a Professor of Computer Engineering at the University of Toronto, where he has the privilege of working with talented graduate students on techniques for improving the execution time, energy efficiency, and cost of computer hardware. He has also taught at Northwestern University, USA, the University of Athens, Greece, the Open University of Greece, Greece, and as a visiting professor at École Polytechnique Fédérale de Lausanne, Switzerland. He is the Scientific Director of the NSERC COHESA Research Network, a Senior Fellow of the Vector Institute, a Fellow of the ACM and the IEEE, and a co-founder of Tartan AI.
Computing Systems for Enabling Innovation in Machine Learning Applications: Learning and Adapting Computing Systems
Machine Learning allows computers to "learn", "think", "see", "hear", "read", and "write", and generally interact with the physical world in ways that we usually associate only with people and intelligence. These computing systems can compete with and complement human capabilities, and promise to enhance our ability to discover, learn, and benefit from information sources. The application areas are extensive and cover science, medicine, health, commerce, political planning, and security. However, further innovation in machine learning requires computing systems that can store and process ever-increasing amounts of information. This speech will briefly explain why we need to revisit the design of computing systems. Towards this direction, our team develops hardware/software techniques that can learn and adapt to the application requirements, and that can be used to build computing systems that are more energy efficient, more cost-effective, and more reliable.
Take me back to speakers!