Machine Learning and Quantum Physics: Setting Realistic Expectations

Special Aalto Quantum Physics Seminar (Nanotalo). Speaker: Dr. Peter Wittek (Creative Destruction Lab, University of Toronto)

The growing intersection between machine learning and quantum physics is beginning to repeat patterns familiar from classical machine learning: the gap between theoretical research and the practice is widening. On the theory side, we have outstanding results on generalizing statistical learning theory, we discovered and exploited connections between the representation power of deep learning networks and the methods used in many-body physics, and we have many quantum learning protocols with a promised exponential advantage. On the practical side, we have a handful of basic quantum optimization and sampling routines running on small-scale quantum computers and elementary feedforward architectures can recover dusty results in the study of phase transitions. This talk will give an introduction to the topic and highlight specific problems that are worth addressing to move the field forward.

Peter Wittek obtained his PhD from the National University of Singapore. He is the Academic Director of the Quantum Machine Learning Initiative in the Creative Destruction Lab in the University of Toronto. His research explores the synergies between artificial intelligence, machine learning, quantum information, and high-performance computing. He published the book "Quantum Machine Learning: What Quantum Computing Means to Data Mining", the first monograph on the subject. He has been deeply involved in entrepreneurship and he works on incubating startups that use quantum technologies for machine learning.