The brain is a nonlinear and highly Recurrent Neural Network (RNN). This RNN is surprisingly plastic and supports our astonishing ability to learn and execute complex tasks. However, learning is incredibly complicated due to the brain's nonlinear nature and the obscurity of mechanisms for determining the contribution of each synapse to the output error. This issue is known as the Credit Assignment Problem (CAP) and is a fundamental challenge in neuroscience and Artificial Intelligence (AI). Nevertheless, in the current understanding of cognitive neuroscience, it is widely accepted that a feedback loop systems play an essential role in synaptic plasticity. With this as inspiration, we propose a computational model by combining Neural Networks (NN) and nonlinear optimal control theory. The proposed framework involves a new NN-based actor-critic method which is used to simulate the error feedback loop systems and projections on the NN's synaptic plasticity so as to ensure that the output error is minimized.
翻译:大脑是一个非线性且高度经常出现的神经网络(RNN) 。 这个 RNN是令人惊讶的塑料,支持我们惊人的学习和执行复杂任务的能力。 但是,由于大脑的非线性以及确定每个突触对输出错误所作贡献的机制的模糊性,学习是极其复杂的。 这个问题被称为信用分配问题,是神经科学和人工智能(AI)中的一项基本挑战。 然而,在目前对认知神经科学的理解中,人们普遍认为反馈循环系统在合成可塑性方面起着关键作用。 有了这种启发性,我们建议了一个计算模型,将神经网络和非线性最佳控制理论结合起来。 拟议的框架涉及一个新的基于NNN的行为者-种族方法,用于模拟错误反馈循环系统和对NN的合成合成性塑料的预测,以确保输出错误最小化。</s>