An extension of Billingsley’s uniform boundedness theorem to higher dimensional M-processes. (English) Zbl 0633.60008

Jurečková, Jana; Sen, Pranab Kumar

Kybernetika 23, 382-387 (1987).

Let T be the unit hypercube in \(R^ p\) (i.e., \(T=[0,1]^ p)\), and \(\{X_ n(t)=(X^ 1_ n(t),X^ 2_ n(t),...,X_ n^ p(t)):\) \(t\in T\}^ a \)sequence of multiparameter random processes with values in \(D[0,1]^ p\). Define \[ X^*_ n=\max_{1\leq j\leq p}\sup_{t\in T}| X^ j_ n(t)|. \] When there exist positive constants \(r>1/2\) and \(L_ p\) such that for every \(n\geq 1\) \[ P\{X^*_ n\geq \lambda \}\leq L_ p\lambda^{-2r}\text{ for every } \lambda >0, \] we say that the sequence \(\{X_ n\}\) is uniformly bounded in probability. For robust estimation in general linear models, it may be convenient to consider some general multiparameter M-processes and to exploit their asymptotic linearity results in the study of properties of the derived estimators. In this context, the uniform boundedness in probability is a basic requirement, but the tightness is not.
In this paper the authors give a sufficient condition which assures the uniform boundedness in probability of M-processes with \(r\geq 1\). The condition is less stringent than that one of P. J. Bickel and M. J. Wichura [Ann. Math. Stat. 42, 1656-1670 (1971; Zbl 0265.60011)], which assures the tightness for general multiparameter stochastic processes.

Reviewer: H.Takahata

MSC:

60B10	Convergence of probability measures
60E15	Inequalities; stochastic orderings
62F35	Robustness and adaptive procedures (parametric inference)
62E20	Asymptotic distribution theory in statistics

Keywords:

robust estimation in general linear models; asymptotic linearity results; tightness; uniform boundedness in probability; multiparameter stochastic processes

References:

[1]	P. Bickel, M. J. Wichura: Convergence criteria for multiparameter stochastic processes and some applications. Ann. Math. Statist. 42 (1971), 1656-1670. · Zbl 0265.60011 · doi:10.1214/aoms/1177693164
[2]	P. Billingsley: Convergence of Probability Measures. J. Wiley, New York 1968. · Zbl 0172.21201
[3]	J. Jurečková, P. K. Sen: Invariance principles for some stochastic processes related to M-estimators and their role in sequential statistical inference. Sankhya, Ser. A 43 (1981), 190-210.
[4]	J. Jurečková, P. K. Sen: Sequential procedures based on M-estimators with discontinuous score functions. J. Statist. Plan Infer. 5 (1981), 253-266. · Zbl 0482.62071
[5]	J. Jurečková, P. K. Sen: On adaptive scale-equivariant M-estimators in linear models. Statist. & Dec, Suppl. Issue No. 1 (1984), 31 - 46.
[6]	J. Jurečková, P. K. Sen: A second order asymptotic distributional representation of M-estimators with discontinuous score functions. Ann. Probab. 15 (1987), 814-823. · Zbl 0635.62017 · doi:10.1214/aop/1176992174
[7]	A. H. Welsh: Bahadur representations for robust scale estimators based on regression residuals. Ann. Statist. 14 (1986), 1246-1251. · Zbl 0604.62028 · doi:10.1214/aos/1176350064