A Note on Using State-Dependent Models With a Time-Dependent Variance
扩展了Priestley的状态依赖模型，允许残差方差按移动平均方案随时间变化，以提升某些序列的模型表现。
This research presents an extension of the state-dependent models (SDM) developed by Priestley. For some series, improvement in SDM performance may be achieved by permitting the residual variance to vary through time according to a moving-average scheme.