Department of Electrical Engineering and Computer Science Montefiore Institute, University of Liège, Liège, Belgium.
PLoS One. 2020 Jan 27;15(1):e0227922. doi: 10.1371/journal.pone.0227922. eCollection 2020.
Animals excel at adapting their intentions, attention, and actions to the environment, making them remarkably efficient at interacting with a rich, unpredictable and ever-changing external world, a property that intelligent machines currently lack. Such an adaptation property relies heavily on cellular neuromodulation, the biological mechanism that dynamically controls intrinsic properties of neurons and their response to external stimuli in a context-dependent manner. In this paper, we take inspiration from cellular neuromodulation to construct a new deep neural network architecture that is specifically designed to learn adaptive behaviours. The network adaptation capabilities are tested on navigation benchmarks in a meta-reinforcement learning context and compared with state-of-the-art approaches. Results show that neuromodulation is capable of adapting an agent to different tasks and that neuromodulation-based approaches provide a promising way of improving adaptation of artificial systems.
动物在适应环境方面表现出色，它们能够调整自己的意图、注意力和行为，从而能够高效地与丰富多彩、变幻莫测的外部世界进行交互，而这正是智能机器目前所缺乏的能力。这种适应能力主要依赖于细胞神经调节，这是一种生物学机制，可以动态地以上下文相关的方式控制神经元的内在特性及其对外界刺激的反应。在本文中，我们受到细胞神经调节的启发，构建了一种新的深度神经网络架构，该架构专门用于学习自适应行为。我们在元强化学习环境下的导航基准测试中测试了网络的自适应能力，并与最先进的方法进行了比较。结果表明，神经调节能够使智能体适应不同的任务，并且基于神经调节的方法为提高人工系统的自适应能力提供了一种有前途的途径。
