[finishied] ipi seminar 17:00-18:30, Friday July 21, 2023
知の物理学研究センター / Institute for Physics of Intelligence (ipi)
【Date】
【Speaker】 Hayata YAMASAKI / The University of Tokyo
【Title】
Quantum Ridgelet Transform: Winning Lottery Ticket of Neural Networks with Quantum Computation
【Abstract】
Quantum computation can achieve significant speedups in solving a class of computational problems compared to the best existing algorithms based on conventional classical computation. Quantum machine learning (QML) is an emerging field of research to take advantage of quantum computation in accelerating machine-learning tasks. This talk explains the basics of quantum computation and how quantum computation applies to accelerating machine-learning tasks with theoretical guarantees beyond heuristics, based on a series of works of the speaker.
Ridgelet transform has been a fundamental mathematical tool in the theoretical studies of neural networks. However, the practical applicability of ridgelet transform to conducting learning tasks was limited since its numerical implementation by conventional classical computation requires an exponential runtime exp(O(D)) as data dimension D increases. To address this problem, we develop a quantum ridgelet transform (QRT), which implements the ridgelet transform of a quantum state within a linear runtime O(D) of quantum computation. As an application, we also show that one can use QRT as a fundamental subroutine for QML to efficiently find a sparse trainable subnetwork of large shallow wide neural networks without conducting large-scale optimization of the original network. This application discovers an efficient way in this regime to demonstrate the lottery ticket hypothesis on finding such a sparse trainable neural network. These results open an avenue of QML for accelerating learning tasks with commonly used classical neural networks.
The talk is based on the following papers.
https://arxiv.org/abs/2301.11936
https://arxiv.org/abs/2106.09028
https://arxiv.org/abs/2004.10756
【Venue】
Room 913, 9F, Bldg. #1, Faculty of Science