报告题目：Machine learning to boost quantum algorithms and quantum chemistry: Quantum chemistry in AI & NISQ era
主 讲人：程立雪 腾讯量子实验室
时 间：2023年05月05日10:00
腾讯会议：468-964-854
主办单位：物理与光电工程学院
欢迎各位老师、同学届时前往！
报告摘要：
How to solve quantum problems equation efficiently and accurately are ubiquitous and computationally hard. Currently, two most promising paths to overcome the exponential wall in the many-body Schrödinger equation is quantum computing and machine learning.
We will focus on the application of ML to boost the performances of quantum algorithms and offer potential computational advantages. Quantum approximate optimization algorithm (QAOA), one of the most representative quantum-classical hybrid algorithms, is designed to solve certain combinatorial optimization problems. To enhance the performance of QAOA, we design double adaptive-region Bayesian optimization (DARBO), an adaptive classical optimizer for QAOA. Our experimental results demonstrate that the algorithm greatly outperforms conventional gradient-based and gradient-free optimizers in terms of speed, accuracy, and stability. We also address the issues of measurement efficiency and the suppression of quantum noise by successfully conducting the full optimization loop on the superconducting quantum processor. We will also briefly introduce our software, TenCirChem on using quantum computing to better deal with quantum chemistry problem. We believe that the combo of AI- and Quantum-driven approaches would open the new era in quantum chemistry and pave the way to realize the efficient quantum chemistry simulation via AI-boosted quantum computing.
报告人简介：
程立雪，腾讯量子实验室的高级研究科学家，将于2023年7月加入柏林Microsoft Research AI4Sci团队。2016年获得威斯康星大学麦迪逊分校化学、数学、生物化学和分子生物学双专业理学学士学位，辅修计算机科学。2017年进入加州理工学院攻读硕士学位并获得研究生奖学金。2022年获得加州理工学院理论化学博士学位，师从 Thomas F. Miller III 教授。