With the advent of real-world quantum computing, the idea that parametrized quantum computations can be used as hypothesis families in a quantum-classical machine learning system is gaining increasing traction. Such hybrid systems have already shown the potential to tackle real-world tasks in supervised and generative learning, and recent works have established their provable advantages in special artificial tasks. Yet, in the case of reinforcement learning, which is arguably most challenging and where learning boosts would be extremely valuable, no proposal has been successful in solving even standard benchmarking tasks, nor in showing a theoretical learning advantage over classical algorithms. In this work, we achieve both. We propose a hybrid quantum-classical reinforcement learning model using very few qubits, which we show can be effectively trained to solve several standard benchmarking environments. Moreover, we demonstrate, and formally prove, the ability of parametrized quantum circuits to solve certain learning tasks that are intractable to classical models, including current state-of-art deep neural networks, under the widely-believed classical hardness of the discrete logarithm problem.
With the advent of real-world quantum computing, the idea that parametrized quantum computations can be used as hypothesis families in a quantum-classical machine learning system is gaining increasing traction. Such hybrid systems have already shown the potential to tackle real-world tasks in supervised and generative learning, and recent works have established their provable advantages in special artificial tasks. Yet, in the case of reinforcement learning, which is arguably most challenging and where learning boosts would be extremely valuable, no proposal has been successful in solving even standard benchmarking tasks, nor in showing a theoretical learning advantage over classical algorithms. In this work, we achieve both. We propose a hybrid quantum-classical reinforcement learning model using very few qubits, which we show can be effectively trained to solve several standard benchmarking environments. Moreover, we demonstrate, and formally prove, the ability of parametrized quantum circuits to solve certain learning tasks that are intractable to classical models, including current state-of-art deep neural networks, under the widely-believed classical hardness of the discrete logarithm problem.
Variational quantum circuits have recently gained popularity as quantum machine learning models. While considerable effort has been invested to train them in supervised and unsupervised learning settings, relatively little attention has been given to their potential use in reinforcement learning. In this work, we leverage the understanding of quantum policy gradient algorithms in a number of ways. First, we investigate how to construct and train reinforcement learning policies based on variational quantum circuits. We propose several designs for quantum policies, provide their learning algorithms, and test their performance on classical benchmarking environments. Second, we show the existence of task environments with a provable separation in performance between quantum learning agents and any polynomial-time classical learner, conditioned on the widely-believed classical hardness of the discrete logarithm problem. We also consider more natural settings, in which we show an empirical quantum advantage of our quantum policies over standard neural-network policies. Our results constitute a first step towards establishing a practical near-term quantum advantage in a reinforcement learning setting. Additionally, we believe that some of our design choices for variational quantum policies may also be beneficial to other models based on variational quantum circuits, such as quantum classifiers and quantum regression models.
The main objective of this deliverable is the theoretical development of quantum machine learning machinery for reinforcement learning, which is near-term-device friendly, yet sufficiently general to be applicable to the task of inventory management. This deliverable documents this achievement. With the advent of real-world quantum computing, the idea that parametrized quantum computations can be used as hypothesis families in a quantum-classical machine learning system is gaining increasing traction. Such hybrid quantum-classical systems, which use a classical loop over a parameterized quantum circuit, have already shown the potential to tackle real-world tasks in supervised and generative learning, and recent works have established their provable advantages in special artificial tasks. Yet, in the case of reinforcement learning, which is arguably most challenging and where learning boosts would be extremely valuable, no proposal has been successful in solving even standard benchmarking tasks, nor in showing a theoretical learning advantage over classical algorithms. In this work, we achieve both. We propose a hybrid quantum-classical reinforcement learning model using very few qubits, which we show can be effectively trained to solve several standard benchmarking environments. Moreover, we demonstrate, and formally prove, the ability of parametrized quantum circuits to solve certain learning tasks that are intractable to classical models, including current state-of-art deep neural networks, under the widely-believed classical hardness of the discrete logarithm problem. Our approach is general, and has shown applicable to both discrete and continuous state domains, with discrete actions, and is thus sufficiently general for application to the problem of inventory management. Furthermore, it is built around parametrized circuits, and is thus within the accepted paradigm for near-term-friendly proposals
Research on quantum computing is still in its infancy, but it has a lot of potential uses. One topic with potential is machine learning, namely in the field of reinforcement learning. This work examines the integration of parametrized quantum circuits (PQC) into reinforcement learning (RL) algorithms, assessing the potential of quantum-enhanced models to solve classical RL tasks. It closely follows the example found on the TensorFlow website. This paper reviews applications of quantum reinforcement learning (QRL). We examine PQCs in a standard RL scenario, the CartPole-v1 environment from Gym, using TensorFlow Quantum and Cirq, to evaluate the relative performance of quantum versus conventional models. In comparison to conventional deep neural network (ONN) models, PQCs show slower convergence and higher processing needs, even if they are still able to learn the task and perform competitively. After they are fully trained, the quantum models show unique difficulties during the early training stages and reach a performance stability level like classical methods. This study sheds light on the present constraints as well as possible uses of quantum computing in reinforcement learning, particularly in situations with intricate, high-dimensional settings that prove difficult for classical computers to handle effectively. As we look to the future, we suggest that investigating hybrid quantum-classical algorithms, developing quantum hardware, and using quantum RL for increasingly difficult tasks are essential first steps. The study presents findings from both a classical reinforcement learning algorithm and a quantum integrated reinforcement learning algorithm. To provide a reliable comparison between quantum reinforcement algorithms and their classical equivalents, further work remains. This work lays the groundwork for future advances in the field by investigating the viability and use of quantum algorithms in reinforcement learning, even if it is not particularly unique. The purpose of this work is to help newcomers to this emerging field of study.
Quantum computing has a superior advantage in tackling specific problems, such as integer factorization and Simon's problem. For more general tasks in machine learning, by applying variational quantum circuits, more and more quantum algorithms have been proposed recently, especially in supervised learning and unsupervised learning. However, little work has been done in reinforcement learning, arguably more important and challenging. Previous work in quantum reinforcement learning mainly focuses on discrete control tasks where the action space is discrete. In this work, we develop a quantum reinforcement learning algorithm based on soft actor-critic -- one of the state-of-the-art methods for continuous control. Specifically, we use a hybrid quantum-classical policy network consisting of a variational quantum circuit and a classical artificial neural network. Tested in a standard reinforcement learning benchmark, we show that this quantum version of soft actor-critic is comparable with the original soft actor-critic, using much less adjustable parameters. Furthermore, we analyze the effect of different hyper-parameters and policy network architectures, pointing out the importance of architecture design for quantum reinforcement learning.
Understanding the power and limitations of quantum access to data in machine learning tasks is primordial to assess the potential of quantum computing in artificial intelligence. Previous works have already shown that speed-ups in learning are possible when given quantum access to reinforcement learning environments. Yet, the applicability of quantum algorithms in this setting remains very limited, notably in environments with large state and action spaces. In this work, we design quantum algorithms to train state-of-the-art reinforcement learning policies by exploiting quantum interactions with an environment. However, these algorithms only offer full quadratic speed-ups in sample complexity over their classical analogs when the trained policies satisfy some regularity conditions. Interestingly, we find that reinforcement learning policies derived from parametrized quantum circuits are well-behaved with respect to these conditions, which showcases the benefit of a fully-quantum reinforcement learning framework.
With the rapid advancement of quantum computing, hybrid quantum–classical machine learning has shown numerous potential applications at the current stage, with expectations of being achievable in the noisy intermediate-scale quantum (NISQ) era. Quantum reinforcement learning, as an indispensable study, has recently demonstrated its ability to solve standard benchmark environments with formally provable theoretical advantages over classical counterparts. However, despite the progress of quantum processors and the emergence of quantum computing clouds, implementing quantum reinforcement learning algorithms utilizing parameterized quantum circuits (PQCs) on NISQ devices remains infrequent. In this work, we take the first step towards executing benchmark quantum reinforcement problems on real devices equipped with at most 136 qubits on the BAQIS Quafu quantum computing cloud. The experimental results demonstrate that the policy agents can successfully accomplish objectives under modified conditions in both the training and inference phases. Moreover, we design hardware-efficient PQC architectures in the quantum model using a multi-objective evolutionary algorithm and develop a learning algorithm that is adaptable to quantum devices. We hope that the Quafu-RL can be a guiding example to show how to realize machine learning tasks by taking advantage of quantum computers on the quantum cloud platform.