Abstract:In recent years, advancements in brain science and neuroscience have significantly influenced the field of computer science, particularly in the domain of reinforcement learning (RL). Drawing insights from neurobiology and neuropsychology, researchers have leveraged these findings to develop novel mechanisms for understanding intelligent decision-making processes in the brain. Concurrently, the emergence of quantum computing has opened new frontiers in artificial intelligence, leading to the development of quantum machine learning (QML). This study introduces a novel model that integrates quantum spiking neural networks (QSNN) and quantum long short-term memory (QLSTM) architectures, inspired by the complex workings of the human brain. Specifically designed for reinforcement learning tasks in energy-efficient environments, our approach progresses through two distinct stages mirroring sensory and memory systems. In the initial stage, analogous to the brain's hypothalamus, low-level information is extracted to emulate sensory data processing patterns. Subsequently, resembling the hippocampus, this information is processed at a higher level, capturing and memorizing correlated patterns. We conducted a comparative analysis of our model against existing quantum models, including quantum neural networks (QNNs), QLSTM, QSNN and their classical counterparts, elucidating its unique contributions. Through empirical results, we demonstrated the effectiveness of utilizing quantum models inspired by the brain, which outperform the classical approaches and other quantum models in optimizing energy use case. Specifically, in terms of average, best and worst total reward, test reward, robustness, and learning curve.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to develop a new reinforcement - learning framework by combining quantum computing and brain - inspired models in the context of energy - efficiency optimization scenarios. Specifically, the authors introduce a new model that integrates the Quantum Spiking Neural Network (QSNN) and the Quantum Long - Short - Term Memory Network (QLSTM) architectures, aiming to simulate the complex working principles of the human brain and being especially suitable for reinforcement - learning tasks that require efficient energy utilization. The model is designed with two stages to process information, similar to the brain's sensory system and memory system: first, extract low - level information like the hypothalamus, simulating the sensory data - processing mode; second, process this information at a higher level like the hippocampus, capturing and remembering relevant patterns. In this way, the model can not only effectively filter noise and infrequent events, but also strengthen information with strong spatio - temporal correlations, thereby achieving a smooth transition from short - - term to long - - term memory, while retaining time and space information and enabling the system to continuously learn new knowledge without forgetting previously acquired knowledge. The paper also explores the performance of this model under different architectures, especially compared with existing quantum models (such as the Quantum Neural Network QNN, QLSTM, QSNN and their classical counterparts), demonstrating its advantages in optimizing energy - use cases, especially in terms of average, best and worst total rewards, test rewards, robustness and learning curves. In summary, this research aims to provide a more energy - efficient and efficient solution for reinforcement - learning through brain - inspired quantum models.

Brain-Inspired Agents for Quantum Reinforcement Learning

Quantum agents in the Gym: a variational quantum algorithm for deep Q-learning

Efficient quantum recurrent reinforcement learning via quantum reservoir computing

Reinforcement Learning with Neural Networks for Quantum Feedback

Quantum Machine Learning Architecture Search via Deep Reinforcement Learning

A Quantum Leaky Integrate-and-Fire Spiking Neuron and Network

Quantum enhancements for deep reinforcement learning in large spaces

Experimental quantum speed-up in reinforcement learning agents

An Introduction to Quantum Reinforcement Learning (QRL)

Neural Predictor Based Quantum Architecture Search

Trends in quantum reinforcement learning: State‐of‐the‐arts and the road ahead

A reinforcement learning approach for quantum state engineering

Variational Quantum Circuits for Deep Reinforcement Learning

Variational Quantum Circuit Design for Quantum Reinforcement Learning on Continuous Environments

Markovian Quantum Neuroevolution for Machine Learning

On the convergence of projective-simulation-based reinforcement learning in Markov decision processes

Quantum Reinforcement Learning in Non-Abelian Environments: Unveiling Novel Formulations and Quantum Advantage Exploration

Challenges for Reinforcement Learning in Quantum Circuit Design

Reinforcement learning-based architecture search for quantum machine learning

Quantum Multi-Agent Reinforcement Learning for Aerial Ad-hoc Networks

Quantum noise modeling through Reinforcement Learning