In this talk, I will report two recent progresses concerning tensor network (TN). Firstly, the quantum entanglement simulation for simulating the ground states and thermodynamics of infinite-size quantum lattice models will be introduced. TN is utilized to construct the few-body Hamiltonians of only O(10) sites to efficiently access the many-body physics including criticality and topological properties. Secondly, I will introduce two novel machine learning schemes, where TN is revealed as a powerful mathematic model that is competitive to the conventional ML such as neural network. Moreover, TN provides a natural way of implementing machine learning on quantum computers. An entanglement-guided TN scheme is introduced that allows for quantum machine learning with a handful of computing qubits.

冉仕举, 2010年本科毕业于北京师范大学物理学系; 2015年博士毕业于中国科学院大学物理学院; 2015至2018年于西班牙光子科学研究所从事博士后研究, 并于2017年获Fundacio-Catalunya独立博士后研究员fellowship; 2018年入职首都师范大学物理系。研究方向是强关联系统的数值计算与量子模拟、量子机器学习模型与算法, 主要使用的理论工具是张量网络。张量网络是一种起源于量子信息科学的强大的数值工具, 其不但可用于高效处理量子多体系统, 最近还被用于发展量子多体态空间的机器学习模型, 实现基于量子理论的人工智能。