Probabilistic Properties Research Articles

Bayesian Empirical Likelihood Regression for Semiparametric Estimation of Optimal Dynamic Treatment Regimes.

We propose a semiparametric approach to Bayesian modeling of dynamic treatment regimes that is built on a Bayesian likelihood-based regression estimation framework. Methods based on this framework exhibit a probabilistic coherence property that leads to accurate estimation of the optimal dynamic treatment regime. Unlike most Bayesian estimation methods, our proposed method avoids strong distributional assumptions for the intermediate and final outcomes by utilizing empirical likelihoods. Our proposed method allows for either linear, or more flexible forms of mean functions for the stagewise outcomes. A variational Bayes approximation is used for computation to avoid common pitfalls associated with Markov Chain Monte Carlo approaches coupled with empirical likelihood. Through simulations and analysis of the STAR*D sequential randomized trial data, our proposed method demonstrates superior accuracy over Q-learning and parametric Bayesian likelihood-based regression estimation, particularly when the parametric assumptions of regression error distributions may be potentially violated.

Threshold integer-valued autoregressive model with serially dependent innovation

ABSTRACT Nonlinearity in count time series is commonly encountered in practice. To better explain the nonlinear phenomena in count time series, this article introduces a new threshold INAR(1) model using the idea of serially dependent innovation. The proposed model contains some existing INAR(1) models as special cases and establishes a connection between the threshold INAR(1) model and the threshold INARCH(1) model. The basic probabilistic and statistical properties of the new model are investigated. Model parameters, including the threshold variable, are estimated by the conditional least squares, modified quasi-likelihood and conditional maximum likelihood methods. Asymptotic properties and numerical results of the estimates are also studied. An application to the monthly trading volume of stock B in Shanghai Stock Exchange is conducted to show the practicability of the proposed model.

Periodic INAR(1) model with Bell innovations distribution

Abstract In this paper, we introduce a class of periodic integer-valued autoregressive BL-PINAR(1) models with Bell innovations distribution based on the binomial thinning operator. The basic probabilistic and statistical properties of this class are studied. Indeed, the first and the second moment periodically stationary conditions are established. The closed forms of these moments are, under the obtained conditions, derived. Furthermore, the periodic autocovariance structure is also considered while providing the closed form of the periodic autocorrelation function. The conditional least squares (CLS), Yule–Walker (YW), weighted conditional least squares (WCLS), and conditional maximum likelihood (CML) methods are applied to estimate the underlying parameters. The asymptotic properties of the CLS and the YW estimators are obtained. The performances of these methods are compared through a simulation study. An application on a real data set is provided.

On some mixtures of the Kies distribution

The purpose of this paper is to explore some mixtures, discrete and continuous, based on the Kies distribution. Some conditions for convergence are established. We study the probabilistic properties of these mixtures. Special attention is taken to the so-called Hausdorff saturation. Several models are examined in detail – bimodal, multimodal, and mixtures based on binomial, geometric, exponential, gamma, and beta distributions. We provide some numerical experiments for real-life tasks – one for the Standard and Poor’s 500 financial index and another for unemployment insurance issues. In addition, we check the consistency of the proposed estimator using generated data of different sizes.

Probabilistic analysis of the steady state of weakly perturbed linear oscillators subject to a class of Gaussian inputs

This paper aims to probabilistically study a class of nonlinear oscillator subject to weak perturbations and driven by stationary zero-mean Gaussian stochastic processes. For the sake of generality in the analysis, we assume that the perturbed term is a polynomial of arbitrary degree in the spatial position, that contains, as a particular case, the important case of the Duffing equation. We then take advantage of the so-called stochastic equivalent linearization technique to construct an equivalent linear model so that its behavior consistently approximates, in the mean-square sense, that of the nonlinear oscillator. This approximation allows us to take extensive advantage of the probabilistic properties of the solution of the linear model and its first mean-square derivative to construct reliable approximations of the main statistical moments of the steady state. From this key information, we then apply the principle of maximum entropy to construct approximations of the probability density function of the steady state. We illustrate the superiority of the equivalent linearization technique over the perturbation method through some examples.

Entropy-extreme model for predicting the development of cyber epidemics at early stages

The approaches used in biomedicine to analyze epidemics take into account features such as exponential growth in the early stages, slowdown in dynamics upon saturation, time delays in spread, segmented spread, evolutionary adaptations of the pathogen, and preventive measures based on universal communication protocols. All these characteristics are also present in modern cyber epidemics. Therefore, adapting effective biomedical approaches to epidemic analysis for the investigation of the development of cyber epidemics is a promising scientific research task. The article is dedicated to researching the problem of predicting the development of cyber epidemics at early stages. In such conditions, the available data is scarce, incomplete, and distorted. This situation makes it impossible to use artificial intelligence models for prediction. Therefore, the authors propose an entropy-extreme model, defined within the machine learning paradigm, to address this problem. The model is based on estimating the probability distributions of its controllable parameters from input data, taking into account the variability characteristic of the last ones. The entropy-extreme instance, identified from a set of such distributions, indicates the most uncertain (most negative) trajectory of the investigated process. Numerical methods are used to analyze the generated set of investigated process development trajectories based on the assessments of probability distributions of the controllable parameters and the variability characteristic. The result of the analysis includes characteristic predictive trajectories such as the average and median trajectories from the set, as well as the trajectory corresponding to the standard deviation area of the parameters’ values. Experiments with real data on the infection of Windows-operated devices by various categories of malware showed that the proposed model outperforms the classical competitor (least squares method) in predicting the development of cyber epidemics near the extremum of the time series representing the deployment of such a process over time. Moreover, the proposed model can be applied without any prior hypotheses regarding the probabilistic properties of the available data.

A novel time-series probabilistic forecasting method for multi-energy loads

Due to the strong nonlinearity, stochasticity, and high coupling of multi-energy loads, this paper proposes a time-series probabilistic forecasting method based on radial basis function network-based autoregressive with exogenous inputs (RBF-ARX) and Gaussian mixture model (GMM). First, due to the nature of the nonlinear coupling between multi-energy loads, the degree of correlation is quantitatively evaluated by introducing the maximum information coefficient with the optimal feature method. Second, the Fourier transform method is used to categorize the load data into deterministic and stochastic components. For the deterministic component, the RBF-ARX method is used for forecasting, which can realize multiple inputs and multiple outputs. It only requires less training data with faster computation speed than other benchmark methods. For the stochastic component, a GMM with an improved Markov Monte Carlo sampling is used to generate the time-series stochastic component, which can accurately characterize the probabilistic properties. This paper compares the forecasting accuracy and computation time under 4 seasons using the proposed model with benchmark models. Results show that the proposed model can improve forecasting accuracy of multi-energy loads by at least 50 % and the computational efficiency can be improved by at least 30 %, which means a superior performance.

Marshall–Olkin Bivariate Weibull Model with Modified Singularity (MOBW-μ): A Study of Its Properties and Correlation Structure

We propose the “Marshall–Olkin Bivariate Weibull Model with Modified Singularity MOBW-μ”, which focuses on bivariate distributions essential for reliability and survival analyses. Distributions such as the Marshall–Olkin bivariate exponential (MOBE) and the Marshall–Olkin bivariate Weibull (MOBW) are discussed. The MOBW-μ model is introduced, which incorporates a lag parameter μ in the singular part, and probabilistic properties such as the joint survival function, marginal density functions, and the bivariate hazard rate function are explored. In addition, aspects such as the correlation structure and survival copulation are addressed and we show that the correlation of the MOBW-μ is always lower than that of its copula, regardless of the parameters. The latter result implies that the MOBW-μ does not have the Lancaster’s phenomenon that explains that any nonlinear transformation of variables decreases the correlation in absolute value. The article concludes by presenting a robust theoretical framework applicable to various disciplines.

Fatigue Life Data Fusion Method of Different Stress Ratios Based on Strain Energy Density.

To accurately evaluate the probabilistic characteristics of the fatigue properties of materials with small sample data under different stress ratios, a data fusion method for torsional fatigue life under different stress ratios is proposed based on the energy method. A finite element numerical modeling method is used to calculate the fatigue strain energy density during fatigue damage. Torsional fatigue tests under different stresses and stress ratios are carried out to obtain a database for research. Based on the test data, the Wt-Nf curves under a single stress ratio and different stress ratios are calculated. The reliability of the models is illustrated by the scatter band diagram. More than 85% of points are within ±2 scatter bands, indicating that the fatigue life under different stress ratios can be represented by the same Wt-Nf curve. Furthermore, P-Wt-Nf prediction models are established to consider the probability characteristics. According to the homogeneity of the Wt-Nf model under different stress ratios, we can fuse the fatigue life data under different stress ratios and different strain energy densities. This data fusion method can expand the small sample test data and reduce the dispersion of the test data between different stress ratios. Compared with the pre-fusion data, the standard deviations of the post-fusion data are reduced by a maximum of 21.5% for the smooth specimens and 38.5% for the notched specimens. And more accurate P-Wt-Nf curves can be obtained to respond to the probabilistic properties of the data.

Probabilistic analysis of homogenized elastic property for resin products fabricated by additive manufacturing based on three-dimensional random field modeling of microstructure

Additive manufacturing (AM) techniques have been used in several fields of science and industry, and fabrication techniques are being updated. For this fact, especially, for industrial use, mechanical property evaluation methodologies for AM products and standards for product quality assessment should also be well established. In this paper, a probabilistic evaluation of the homogenized elastic properties of a resin product fabricated by a material extrusion-based AM technique is attempted by considering the randomness of both material and microscopic geometrical quantities. This AM method fabricates a resin structure by piling up melted resin, and to decrease consumed material and influence of thermal deformation, the inner structure of the fabricated products will include many pores and its geometry is difficult to be well controlled. From this fact, the products will be regarded as a heterogeneous material with complex random microstructure. This will cause difficulty in the evaluation of its apparent material properties and therefore a probabilistic homogenization analysis is attempted for their quantitative estimation in this study. In particular, to investigate probabilistic properties of microscopic geometry, a random field modeling technique is employed for the evaluation of autocorrelation of the microscopic geometrical parameter, and the results of the autocorrelation identified by experimental observation are introduced to the probabilistic homogenization analysis. The two-dimensional or three-dimensional random field modeling is attempted, and the effectiveness of this approach is investigated by comparing it with the experimental result.

Bivariate first-order random coefficient integer-valued autoregressive processes based on modified negative binomial operator

In this paper, a new bivariate random coefficient integer-valued autoregressive process based on modified negative binomial operator with dependent innovations is proposed. Basic probabilistic and statistical properties of this model are derived. To estimate unknown parameters, Yule-Walker, conditional least squares and conditional maximum likelihood methods are considered and evaluated by Monte Carlo simulations. Asymptotic properties of the estimators are derived. Moreover, coherent forecasting and possible extension of the proposed model is provided. Finally, the proposed model is applied to the monthly crime datasets and compared with other models.

Population Dependent Two-Sex Branching Process with Random Mating and Overlapping Generations

Abstract This research focuses on stochastic modeling of the evolution over time of biological populations through branching processes. We introduce a new class of discrete time two-sex branching processes with random mating and overlapping generations. Mating and reproduction are considered to be influenced by the numbers of females and males existing in the population. This evolution over time with generational overlap is a novel research in two-sex branching process literature. We study general probabilistic properties and establish some limiting results of biological interest.

Shift-invariant homogeneous classes of random fields

Given an Rd-valued random field (rf) Z(t),t∈T and an α-homogeneous mapping κ we define the corresponding equivalent class of rf's (denoted by Cκ[Z]) which include representers of the same tail measure νZ. When T is an additive group, tractable equivalent classes of interest are the shift-invariant ones, which contain in particular all independent random shifts of Z. This contribution is mainly concerned with the investigation of the probabilistic properties of shift-invariant Cκ[Z]'s. Important objects introduced in our setting are tail and spectral tail rf's. Further, the class of universal maps U acting on elements of Cκ[Z] turns out to be crucial for properties of functionals of Z. Applications of our findings concern max-stable and symmetric α-stable rf's, their maximal indices as well as their random shift-representations.

The Tail Mean–Variance optimal capital allocation under the extended skew-elliptical distribution

In this paper we introduce a novel Tail Mean-Variance (TMV) model intended to optimize capital allocation in financial decision-making. The TMV model has many intriguing properties, such as considering the variability and tail risk of the loss function simultaneously. We specifically consider the multivariate extended skew-elliptical (ESE) distributions, which are widely applicable in financial and insurance data modeling. Additionally, we explore the probabilistic properties of the multivariate extended skew-elliptical (ESE) distributions and present explicit formulas for the TMV model within this distribution. Finally, a numerical example is provided to illustrate the results.

Semantic analysis of two-syllabic terms of metrology. Part 2: Risk in measurements and calculations

Prior to the requirement for testing and calibration laboratories to take into account the risk of statistical assumptions, false positive and false negative decisions in international documents, it was found that the methodology of the “Guide to Expressing Measurement Uncertainty” based on the Bayes approach and the Monte Carlo method for calculating probabilistic risk characteristics is not applicable. A later draft revision of the “Manual on the Expression of Measurement Uncertainty” attempted to shift the interpretation of measurement uncertainty from the scattering parameter to the probability distribution. An attempt to contribute to solving the problem of definitional uncertainty in the International Dictionary of Basic and Basic Terms of Metrology was also unsuccessful. In the new version of the dictionary of general statistical terms and probability theory terms, the term measurement uncertainty is excluded, and one of the notes states that “the probability distribution fully describes the probabilistic properties of the uncertainty of the result”. However, due to the new requirements for risk calculations, international documents were urgently put into effect without radical assessments of the inapplicability of the Bayes approach and the Monte Carlo method, the disadvantages were renamed limitations, but there are no specific instructions for calculating risks. Based on the experience of the compositional approach to estimating accuracy, a procedure based on the convolution of probability distributions in the form of a modified reversal formula is recommended, which allows taking into account the definitive uncertainty in the moment approach. It is established that the method of accounting for the definitive uncertainty by convolution of uniform distributions is practically suggested in the text of the “Manual on the expression of measurement uncertainty”, but not used.

An improved analytical solution to outcrossing rate for scalar nonstationary and non-gaussian processes

For time-dependent reliability methods, there is a lack of analytical solutions to outcrossing rates of nonstationary and non-Gaussian processes. This paper aims to propose an analytical method to determine this type of outcrossing rate with improved accuracy and efficiency. The novelty of this proposed method is that it can consider the probabilistic properties of non-Gaussian processes by combining a two-component parallel system model with higher-order moments-based reliability indexes. The main contributions of this method are: (1) compared with PHI2+ method, it is more accurate for the calculation of the outcrossing rate of nonstationary and non-Gaussian processes with nonlinear limit state functions; (2) compared with MPHI2 method, it is analytical and not only applicable for the nonstationary stochastic processes but also insensitive to time increments; and (3) it is not only more accurate than the existing methods for nonstationary and non-Gaussian processes but also more computation efficient than the Monte Carlo simulation method. The proposed method has shown its advantages for practical structures with neither Gaussian processes nor linear limit state functions, which are beneficial for both researchers and asset managers to evaluate the time-dependent reliability of structures accurately and to develop risk-informed maintenance schemes with a view to prolonging their service life.

Distinguishing between fractional Brownian motion with random and constant Hurst exponent using sample autocovariance-based statistics.

Fractional Brownian motion (FBM) is a canonical model for describing dynamics in various complex systems. It is characterized by the Hurst exponent, which is responsible for the correlation between FBM increments, its self-similarity property, and anomalous diffusion behavior. However, recent research indicates that the classical model may be insufficient in describing experimental observations when the anomalous diffusion exponent varies from trajectory to trajectory. As a result, modifications of the classical FBM have been considered in the literature, with a natural extension being the FBM with a random Hurst exponent. In this paper, we discuss the problem of distinguishing between two models: (i) FBM with the constant Hurst exponent and (ii) FBM with random Hurst exponent, by analyzing the probabilistic properties of statistics represented by the quadratic forms. These statistics have recently found application in Gaussian processes and have proven to serve as efficient tools for hypothesis testing. Here, we examine two statistics-the sample autocovariance function and the empirical anomaly measure-utilizing the correlation properties of the considered models. Based on these statistics, we introduce a testing procedure to differentiate between the two models. We present analytical and simulation results considering the two-point and beta distributions as exemplary distributions of the random Hurst exponent. Finally, to demonstrate the utility of the presented methodology, we analyze real-world datasets from the financial market and single particle tracking experiment in biological gels.

Wind-resistant reliability assessment of a super-high tower crane considering the randomness and correlation of turbulent spectral parameters

The need for wind-sensitive super-high tower cranes during the construction of high-rise structures has been growing, which makes it critical to assess the wind-resistant reliability of cranes. The accuracy of assessment is dominated by the refined simulation of real stochastic wind fields, which is usually based on the deterministic turbulent spectra including fixed spectral parameters. However, it might cause significant errors to estimate extreme responses of structures. Therefore, a simulation method for stochastic wind fields considering the randomness and correlation of turbulent spectral parameters is proposed in this study. The random functions of correlated spectral parameters are constructed based on several independent elementary random variables. A novel simulation method for stochastic wind fields involving the established random turbulent spectral functions is utilized to generate fluctuating wind samples at the Ma'anshan Yangtze River (MYR) Bridge site. In combination with the probability density evolution method, the stochastic response analysis and reliability assessment of wind-induced vibration of the crane are addressed. Results prove the effectiveness of the proposed approach in depicting probabilistic properties of real wind fields. Meanwhile, the wind-resistant comfort reliability of the tower crane is higher when considering the randomness and correlation of turbulent spectral parameters.

Symmetric physically unclonable functions of the arbiter type

Objectives. The problem of constructing a new class of physically unclonable functions of the arbiter type (APUF) that combines the advantages of both classical and balanced APUF is solved. The relevance of such a study is associated with the active development of physical cryptography. The following goals are pursued in the work: research and analysis of classical APUF, construction of a new mathematical model of APUF and development of a new basic element of APUF.Methods. The methods of synthesis and analysis of digital devices are used, including those based on programmable logic integrated circuits, the basics of Boolean algebra and circuitry.Results. It has been established that classical APUF uses a standard basic element that performs three functions, namely, the function of generating two random variables Generate, the function of choosing a pair of paths Select and the function of switching paths Switch, which are specified by one bit of the challenge. It is shown that the joint use of these functions, on the one hand, makes it possible to achieve high characteristics of the APUF, and on the other hand, leads to the formation of an asymmetric behavior of the APUF. In order to analyze the main characteristics of APUF and their ideal behavior, a new mathematical model of APUF was considered, similar to the model of random coin toss. To implement APUF functioning according to the proposed model, a new basic element was developed. It is shown that the use of the proposed basic element allows to build symmetrical physically unclonable functions (C_APUF), which differ from the classical APUF in that the Generate, Select and Switch functions of the basic element are performed by their independent components and are specified by different bits of challenge.Conclusion. The proposed approach to the construction of symmetrical physically unclonable functions, based on the implementation of the Generate, Select and Switch functions by various components of the base element, has shown its efficiency and promise. The effect of improving the characteristics of similar C_APUF has been experimentally confirmed, and, first of all, a noticeable improvement in their probabilistic properties expressed in equal probability of responses. It seems promising to further develop the ideas of building C_APUF, experimental study of their characteristics, as well as analysis of resistance to various types of attacks, including using machine learning.

On the existence of stationary threshold bilinear processes

This article investigates some statistical and probabilistic properties of general threshold bilinear processes. Sufficient conditions for the existence of a causal, strictly and weak stationary solution for the equation defining a self-exciting threshold superdiagonal bilinear SETBL\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\left( SETBL\\right) $$\\end{document} process are derived. Then it is shown that under well-specified hypotheses the higher-order moments of the SETBL process are finite. As a result, the skewness and kurtosis indexes are explicitly computed. The exact autocorrelation function is derived with an arbitrarily fixed number of regimes. Also, the covariance functions of the process and its powers are evaluated and the second (respectively, higher)-order structure is shown to be similar to that of a linear process. This implies that the considered process admits an ARMA representation. Finally, necessary and sufficient conditions for the invertibility and geometric ergodicity of a SETBL model are established. Some examples illustrate the obtained theoretical results.