Self-consistent signal-to-noise analysis and its application to analogue neural networks with asymmetric connections
Self-consistent signal-to-noise analysis and its application to analogue neural networks with asymmetric connections
复制标题
自洽信噪比分析及其在不对称连接模拟神经网络中的应用
DOI：
10.1088/0305-4470/25/7/017
复制
发表时间：
1992
期刊：
影响因子：
--
通讯作者：
T. Fukai
中科院分区：
文献类型：
--
作者：
M. Shiino;T. Fukai
本文提出了一种新的系统方法，用于分析具有一般输入输出关系的模拟神经网络的存储容量。它是基于自洽的信号-噪声分析，其中在本地字段中的信号部分的重整化适当地执行。本配方的一个显着特点，在对称模拟网络的情况下，产生相同的结果，通过复制计算获得的，是处理模拟神经元的非对称网络的能力。该理论适用于非对称网络，其中每个神经元加载有偏的模式，而一些神经元只分配给扩展抑制性突触耦合免费学习。
A new systematic method is proposed for the analysis of the storage capacity of analogue neural networks with general input-output relations. It is based on the self-consistent signal-to-noise analysis in which renormalization of the signal part in the local field is properly performed. A remarkable feature of the present recipe, which in the case of symmetric analogue networks yields the same result as obtained by replica calculations, is the capability of dealing with asymmetric networks of analogue neurons. The theory is applied for the asymmetric network in which each neuron is loaded with biased patterns while some neurons are assigned only to extend inhibitory synaptic couplings free of learning.