Hodge-Kodaira decomposition of evolving neural networks
Hodge-Kodaira decomposition of evolving neural networks
复制标题
DOI：
10.1016/j.neunet.2014.05.021
复制
发表时间：
2015-02-01
期刊：
影响因子：
7.8
通讯作者：
Aoki, Takaaki
中科院分区：
文献类型：
--
作者：
Miura, Keiji;Aoki, Takaaki
尽管仔细检查神经网络的循环结构对于阐明大脑功能非常重要，但传统方法往往难以系统地表征网络内的全局循环。在这里，我们应用Hodge-Kodaira分解，一种拓扑方法，一个不断发展的神经网络模型，以表征其循环结构。通过参数化地控制学习规则，我们发现具有STDP规则的模型具有最多的循环，该STDP规则倾向于形成与因果触发顺序一致的路径。此外，通过计算网络中的全局环的数量，我们发现了混沌区域内的不均匀性，这通常被认为是棘手的。(C)2014爱思唯尔有限公司版权所有。
Although it is very important to scrutinize recurrent structures of neural networks for elucidating brain functions, conventional methods often have difficulty in characterizing global loops within a network systematically. Here we applied the Hodge-Kodaira decomposition, a topological method, to an evolving neural network model in order to characterize its loop structure. By controlling a learning rule parametrically, we found that a model with an STDP-rule, which tends to form paths coincident with causal firing orders, had the most loops. Furthermore, by counting the number of global loops in the network, we detected the inhomogeneity inside the chaotic region, which is usually considered intractable. (C) 2014 Elsevier Ltd. All rights reserved.