Asymptotic results for runs and empirical cumulative entropies (Q473524)

The paper under review deals with sequences of runs and of empirical cumulative entropies. The authors utilize the definition of runs based on differences and ratios of consecutive order statistics as in [\textit{S. Eryilmaz} and \textit{A. Stepanov}, ``Runs in an ordered sequence of random variables'', Metrika 67, No. 3, 299--311 (2008; Zbl 1357.60037); \textit{A. Stepanov}, Stat. Probab. Lett. 81, No. 1, 54--61 (2011; Zbl 1218.62045); Commun. Stat., Theory Methods 40, No. 18, 3252--3268 (2011; Zbl 1225.62068)]. In both cases there are sums of independent Bernoulli distributed random variables, so the authors present large and moderate deviations results by referring to this scheme and then illustrate the connection with random variables studied in the cited papers. The authors refer to the sequence of estimators in [\textit{A. Di Crescenzo} and \textit{M. Longobardi}, J. Stat. Plann. Inference 139, No. 12, 4072--4087 (2009; Zbl 1172.94543)] for the empirical cumulative entropies and, thus, have linear combinations of independent exponentially distributed random variables when proving large and moderate deviations results. Spacings play an important role here when considering both runs and empirical cumulative entropies.