It is well known that when the random errors are iid. with finite variance, the week and the strong consiStency of LS estimate of multiple regression coefficients are equivalent. This note, by constructing a counter-example, shows that this equivalence no longer holds true in case that the random errors possess only the r-th moment with 1≤5 T ＜ 2.
CHEN XIRU The Graduate School at Beijing, University of Science and Technology of China, Beijing 100039, China.
Estimators are presented for the coefficients of the polynomial errors-in-variables (EV) model when replicated observations are taken at some experimental points. These estimators are shown to be strongly consistent under mild conditions.
This paper studies the parameter estimation of one dimensional linear errors-in-variables(EV) models in the case that replicated observations are available in some experimental points.Asymptotic normality is established under mild conditions, and the parameters entering the asymptotic variance are consistently estimated to render the result useable in construction of large-sample confidence regions.
ZHANG SANGUO CHEN XIRUHua Lee-Keng Institue for applied Mathematics and Information Science, Graduate School of ChineseAcademy of Sciences, Beijing 100039, China. Department of Mathematics, Graduate School of Chinese Academy of Sciences,
This paper based on the essay [1], studies in case that replicated observations are available in some experimental points., the parameters estimation of one dimensional linear errors-in-variables (EV) models. Asymptotic normality is established.
In this paper we study the estimation of the rank of the parameter trix(RPM) in a growth curve model in the framework of model selection. Following AIC criterion we propose a new general criterion and obtain a strongly consistent estimate of the RPM. We come to our conclusions under the assumptions of normal population and a general case separately.
ZHAO Lincheng WANG Xuena WU Yaohua (Department of Statistics and Finance, University of Science and Technology of China, Hefei 230026, China)
In this paper, we study a general criterion for estimating the rank of canonical correlation matrix (CCM). Besides the strong consistency, we give the exponential-type bounds under certain conditions on the probability of wrong detection of the rank of CCM. On the basis of this criterion, we give two methods to determine the multiplicities of canonical correlation coefficents. And the strong consistency of them is established.
