scientific article; zbMATH DE number 3911512

From MaRDI portal
Publication:3687535





Consider the linear model \(Y_ i=X^ T_ i\beta +e_ i\), \(i=1,2,...,n,..\). The least squares estimate of the regression coefficient vector \(\beta\) is \[ {\hat \beta}(n)=({\hat \beta}_ 1(n),...,{\hat \beta}_ p(n))=S_ n^{-1}X^ T_ n(Y_ 1,...,Y_ n)^ T,\quad where\quad X_ n=(x_ 1,...,x_ n)^ T,\quad S_ n=X^ T_ nX_ n. \] The present paper proposes a method to improve some previous results. The author also constructs an example when the error sequence with r-th moment is iid., and poses the question whether the r-th moment mean consistency and weak consistency are equivalent? He proves this to be true in the case of Markov sequences, but it is not true in the above case. He generalizes a result concerning the law of large numbers.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3687535)