Cross-fertilizing machine learning and synthetic quantum systems

随着量子力学的初步发展，科学发现随着晶体管、超导、激光等重大技术发明和其他技术的进步而加速。目前，我们正处于第二次量子革命的黎明，利用量子信息处理来产生新技术。量子物理和量子技术中的一个核心概念是多体波函数，它是物理学中最复杂的数学对象之一。值得注意的是，未来量子技术的力量依赖于我们准确控制脆弱量子设备波函数的能力。*作为几个科学领域大趋势的一部分，量子多体物理学家将他们的注意力转向了机器学习的想法，这些想法目前正在化学、材料科学、高能物理等不同领域的科学发现中发生革命性的变化。在最近的一系列发展中，Carrasquilla的研究采取了一个基本观点，即量子系统的状态可以被理解为微观现象的“生成模型”。量子系统的指数复杂性让人想起机器学习中遇到的“维度诅咒”。Carrasquilla研究计划最重要的长期目标是将这一观点扩展到更广泛的机器学习角度来研究量子多体问题，并探索其在量子技术中的应用。这项发现基金提出的目标包括设计受物理量子系统动力学启发的机器学习模型，以及开发量子状态的神经网络表示法。这项研究的应用包括量子态重建，以及以薛定谔方程精确解和近似解的形式在凝聚态物理和量子化学中的应用。*该项目的核心是关于人工智能和深度学习的大量研究内容。这要么是因为量子技术可能会促进人工智能，要么是因为机器学习可能会推动量子设备和算法的设计。归根结底，我的研究战略和长期愿景旨在推动加拿大量子科学、人工智能和新兴量子产业基础研究的发展。这一战略包括培养硕士和博士生，他们将开创将多体物理中的现代理论和数值方法与人工智能技术相结合的方法，在凝聚态物理、量子模拟、材料科学和量子化学等领域产生影响。反过来，这将产生侧重于量子科学的最先进的人工智能研究，以及一支随时准备解决与加拿大新兴量子人工智能行业相关的研究问题的专业劳动力。**

Following the initial development of quantum mechanics, scientific discovery accelerated with major technological inventions such as the transistor, superconductivity, the laser, and other technological advances. Currently, we are at the dawn a second quantum revolution, where quantum information processing is exploited to generate new technology. A central concept in quantum physics and quantum technologies is the many--body wave function, which is one of the most complex mathematical objects in physics. Remarkably, the power of future quantum technologies relies on our ability to accurately control the wave function of brittle quantum devices. ***As part of a large trend in several scientific areas, quantum many-body physicists have turned their attention to machine learning ideas, which are currently revolutionizing scientific discovery in areas as diverse as chemistry, materials science, high energy physics, and other. In a series of recent developments, Carrasquilla's research has taken the fundamental viewpoint that the state of a quantum system, whose exponential complexity is reminiscent of the "curse of dimensionality" encountered in machine learning, can be understood as a "generative model" of microscopic phenomena. The most important long--term goal of Carrasquilla's research program is to expand this viewpoint into a broader machine learning perspective on the quantum many--body problem and to explore its applications in quantum technology. The goals advanced by this Discovery Grant include the design of machine learning models inspired by the dynamics of physical quantum systems, and the development of neural--network representations of quantum states. Applications of this research include quantum state reconstruction, as well as applications in condensed matter physics and quantum chemistry in the form of exact and approximate solutions to the Schrodinger equation. ***At the heart of this program is a substantial research component on artificial intelligence and deep learning. This is either because quantum technology may boost artificial intelligence or because machine learning may propel the design of quantum devices and algorithms. Ultimately, my research strategy and long--term vision are designed to advance the development of Canada's fundamental research on quantum science, artificial intelligence, and the emerging quantum industry. This strategy includes the training of Master's and PhD students who will pioneer methods hybridizing modern theoretical and numerical methods in many-body physics with artificial intelligence techniques, with impact in areas such as condensed matter physics, quantum simulation, materials science, and quantum chemistry. In turn, this will generate state--of--the--art artificial intelligence research with an emphasis on quantum science, and a specialized workforce ready to tackle research problems relevant to the emerging quantum artificial intelligence industry in Canada.**