Showing papers in "Statistics & Probability Letters in 2018"

TL;DR: In this paper, the tail probabilities for sums of independent geometric or exponential variables, possibly with different parameters, are given for different sets of variables, and explicit bounds for their tail probabilities are given.

TL;DR: Residual extropy was proposed to measure residual uncertainty of a random variable and characterization results of this measure were studied in this article, where the proposed measure of order statistics were also discussed.

TL;DR: There is a lack of consideration of interpretability, uncertainty quantification, applications with limited training data, and selection bias in automated methods for complex data analysis.

TL;DR: In this paper, piecewise deterministic Monte Carlo (DMMC) algorithms are implemented in settings where the parameters live on a restricted domain, and shown how they can be implemented in the restricted domain.

TL;DR: A high quality small sample can produce superior inferences to a low quality large sample and sometimes aggregation into small datasets is better than large individual-level data.

TL;DR: A broad review is given of the impact of big data on various aspects of investigation and there is some but not total emphasis on issues in epidemiological research.

TL;DR: The role of statistics regarding some of the issues raised by big data in this new paradigm is discussed and the name of data learning is proposed to describe all the activities that allow to obtain relevant knowledge from this new source of information.

TL;DR: An elementary proof of the fact that a binomial random variable X with parameters n and 0 strictly exceeds its expectation and both probabilities approach 1 ∕ 2 when n p and n ( 1 − p ) tend to infinity.

TL;DR: In this article, a nonparametric test for covariate-adjusted models is proposed, obtained by using the adjusted response and predictors, and the proposed test statistic has the same limit distribution as when the response and predictor are observed directly.

TL;DR: In this paper, the authors propose multivariate bootstrapping techniques as a means for making inferences about the unknown regression coefficient matrix, which are extensions of those developed in Freedman (1981) which are only appropriate for univariate responses.

TL;DR: A selected survey highlights how earlier ideas in high dimensional problems can be adapted in functional setting.

TL;DR: In this paper, the optimal allocation of clusters for a fixed number of periods in cohort stepped wedge cluster randomized trials is studied. And the optimal design turns more clusters into treatment during the second and final periods, and depends on values of correlation parameters.

TL;DR: A simple variable (feature) weight learning strategy for the Gaussian means algorithm which can automatically determine the number of clusters in a dataset as well is proposed.

TL;DR: In this paper, the optimal allocation of active redundancies for weighted k-out-of-n systems is investigated for the case of one, two and multiple active redundancy in the sense of the usual stochastic order when the system is comprised of independent and heterogeneous components having different weights.

TL;DR: This paper proposes regularized expectile regression with SCAD penalty for analyzing heteroscedasticity in high dimension when the error has finite moments and adopts the CCCP (coupling of the concave and convex procedure) algorithm to solve this problem.

TL;DR: The role of statistics in the era of big data was discussed in the 48th Scientific Meeting of the Italian Statistical Society (MSS 2016) as discussed by the authors, with the aim of engaging a larger audience on an issue which promises to change radically our discipline and, more generally, science as we know it.

TL;DR: There should be more room for publishing negative findings in data science, and the need of developing appropriate concepts, methodology and algorithms is highlighted.

TL;DR: In this paper, quantile oriented sensitivity analysis is performed using the Conditional Tail Expectation risk measure, which is used to build estimators for quantile-oriented sensitivity analysis, and the corresponding indices are rewritten using the same approach.

TL;DR: In this article, a generalized regime-switching GARCH model is proposed to capture the dynamic behavior of volatility in financial market and a Monte-Carlo experiment is used to analyze the dynamics of the volatilities and time-dependent probabilities as well as the behaviors of cumulative impulse response functions.

TL;DR: In this article, the authors discuss another tempered version of FBM/FSM, termed tempered fractional Brownian/stable motion of second kind (TFBM II/TFSM II).

TL;DR: Three principles for the statistical analysis of such large data sets that leverage recent methodological and computational advances are provided, which emphasize the need of embedding distributed and parallel computing in the inferential process.

TL;DR: In this paper, a spectrally positive Levy risk process with Parisian implementation delays in dividend payments is introduced, which means that the dividends can only be paid when the surplus of the risk process has stayed continuously above the barrier b for a certain time r ( > 0 ).

TL;DR: It is shown that one of the most important data science models, correlation networks, can play a significant role in the advancements of Fintech developments.

TL;DR: In this article, the mean-square exponential input-to-state stability for a class of delayed impulsive stochastic Cohen-Grossberg neural networks driven by G -Brownian motion was studied.

TL;DR: Yang and Li as discussed by the authors derived an asymptotic expansion for the finite-time ruin probability under the special Farlie-Gumbel-Morgenstern dependence structure and a technical moment condition on the claim-number process.

TL;DR: The role of statistics for Big Data analysis arising from the emerging field of Data-Centric Engineering is explored, using examples related to sensor-instrumented bridges to highlight a number of issues and challenges.

TL;DR: In this article, the singular values of certain triangular random matrices are studied and the squares of the singular value form a biorthogonal ensemble, and with an appropriate change in the distribution of the diagonal elements, they give the biorhogonal Laguerre ensemble.

TL;DR: This work proposes optimal weighting schemes for both mean and covariance estimations for functional data based on local linear smoothing such that the L 2 rate of convergence is minimized.

TL;DR: An application of Universal Kriging to a massive spatial dataset is shown and some perspectives of future work in this field are presented.