摘要
本文以马尔科夫链理论为工具,研究PLN(概率逻辑神经元)网络的定量性质.我们得到的主要结果是,在一给定网络中,给出各状态收敛到稳定状态的概率,平均收敛步数和方差以及一般PLN网络平均收敛步数的上(下)界估计.给出计算机的一个模拟结果,并与理论结论相对比,以验证理论结果的正确性.
In this paper, the main behaviors of PLN (Probabilistic Logic Neuron) networks are investigated quantitatively using Markov chain theory. The main results are the probability that each state in a given network converges to a stable state, the mean and the variance of the number of steps that each state in a given network converges to a set of stablestates, and the upper (lower) bound of the average number of steps that each state in general PLN networks converge. Computer simulation result is given to verify the analysis.
出处
《计算机学报》
EI
CSCD
北大核心
1993年第1期1-12,共12页
Chinese Journal of Computers
关键词
神经元网络
收敛收
概率
逻辑
PLN networks, Markov chain, traversal state, transition state, transition matrix.
