arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2201.13180cs.LG
收藏

基于预测编码的任意图拓扑学习

Learning on Arbitrary Graph Topologies via Predictive Coding

University of Oxford(牛津大学);TU Wien(维也纳工业大学)

Tommaso Salvatori, Luca Pinchetti, Beren Millidge, Yuhang Song, Tianyi Bao, Rafal Bogacz, Thomas Lukasiewicz

摘要:

在标准深度学习中使用反向传播(BP)进行训练包括两个主要步骤:将数据点映射到其预测的前向传递,以及将此预测的误差通过网络传播回来的反向传递。当目标是最小化特定目标函数时,此过程非常有效。但是,它不允许在具有循环或反向连接的网络上进行训练。这是达到类似大脑能力的障碍,因为新皮层中高度复杂的神经连接的异质结构可能是其有效性的基础。在本文中,我们展示了如何预测编码(PC),在大脑皮层的信息处理的理论,可以用来执行推理和学习的任意图形拓扑结构。我们通过实验展示了这种称为PC图的配方如何通过简单地刺激特定的神经元来灵活地执行相同网络的不同任务,并研究图的拓扑结构如何影响最终性能。我们通过与~BP训练的简单基线进行比较得出结论。

英文摘要:

Training with backpropagation (BP) in standard deep learning consists of two main steps: a forward pass that maps a data point to its prediction, and a backward pass that propagates the error of this prediction back through the network. This process is highly effective when the goal is to minimize a specific objective function. However, it does not allow training on networks with cyclic or backward connections. This is an obstacle to reaching brain-like capabilities, as the highly complex heterarchical structure of the neural connections in the neocortex are potentially fundamental for its effectiveness. In this paper, we show how predictive coding (PC), a theory of information processing in the cortex, can be used to perform inference and learning on arbitrary graph topologies. We experimentally show how this formulation, called PC graphs, can be used to flexibly perform different tasks with the same network by simply stimulating specific neurons, and investigate how the topology of the graph influences the final performance. We conclude by comparing against simple baselines trained~with~BP.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑