This paper deals with the estimation of the hidden factor in Dynamic Generalized Factor Analysis via a generalization of Kalman filtering. Asymptotic consistency is discussed and it is shown that the Kalman one-step predictor is not the right tool while the pure filter yields a consistent estimate.
翻译:本文件通过对Kalman过滤法的概括,论述对动态通用系数分析中隐含因素的估计。讨论了Asympt一致性,并表明Kalman单步预测器不是正确工具,而纯过滤器得出一致的估计值。