摘要
用马氏链理论研究函数型随机方差非线性自回归模型平稳分布和矩的存在性,特别是只假设新息序列中的随机变量有密度函数,而不需要有处处为正的密度函数.
In this paper, we discuss existence of the stationary distribution and moment of nonlinear autoregressive model with conditional heteroscedasticity by the Markov theory. Especially, only assume that random variable has density function in the innovation, but not positive everywhere.
出处
《数学物理学报（A辑）》
CSCD
北大核心
2013年第2期260-266,共7页
Acta Mathematica Scientia
关键词
马氏链
非线性时间序列
一致可数可加
Markov chains
Nonlinear autoregressive
Uniform countable additivity.
