This paper aims to enhance the accuracy and reliability of ship identification processes by preprocessing noise and outliers in the collected data. We propose a preprocessing method using the Expectation-Maximization ...
详细信息
This paper aims to enhance the accuracy and reliability of ship identification processes by preprocessing noise and outliers in the collected data. We propose a preprocessing method using the Expectation-Maximization (em) algorithm and the Extended Kalman Filter (EKF) applied to ship trajectory and motion data, demonstrating significant advancements over existing methodologies. We introduce a novel application of one-dimensional Student's t-distributions with independent parameters by Q and R (measured noise covariance), enabling more precise modeling of sensor noise and outliers, thus improving data processing robustness. Our algorithm also incorporates adaptive learning capabilities, allowing adjustments to various environmental conditions, which enhances noise and outlier handling and broadens applicability across real-world scenarios. Experimental results indicate that our methods significantly improve data accuracy and ship identification reliability, underscoring their practical significance. These techniques provide robust technical support for accurate ship motion information and have considerable potential for maritime identification applications.
A new robust stochastic volatility (SV) model having Student-t marginals is proposed. Our process is defined through a linear normal regression model driven by a latent gamma process that controls temporal dependence....
详细信息
A new robust stochastic volatility (SV) model having Student-t marginals is proposed. Our process is defined through a linear normal regression model driven by a latent gamma process that controls temporal dependence. This gamma process is strategically chosen to enable us to find an explicit expression for the pairwise joint density function of the Student-t response process. With this at hand, we propose a composite likelihood (CL) based inference for our model, which can be straightforwardly implemented with a low computational cost. This is a remarkable feature of our Student-t process over existing SV models in the literature that involve computationally heavy algorithms for estimating parameters. Aiming at a precise estimation of the parameters related to the latent process, we propose a CL expectation-maximization algorithm and discuss a bootstrap approach to obtain standard errors. The finite-sample performance of our CL methods is assessed through Monte Carlo simulations. The methodology is motivated by an empirical application in the financial market. We analyze the relationship, across multiple time periods, between various US sector Exchange-Traded Funds returns and individual companies' stock price returns based on our novel Student-t model. This relationship is further utilized in selecting optimal financial portfolios. Generalizations of the Student-t SV model are also proposed.
The theory of multilevel hierarchical data Expectation Maximization (em)-algorithm is introduced via discrete time Markov chain (DTMC) epidemic models. A general model for a multilevel hierarchical discrete data is de...
详细信息
The theory of multilevel hierarchical data Expectation Maximization (em)-algorithm is introduced via discrete time Markov chain (DTMC) epidemic models. A general model for a multilevel hierarchical discrete data is derived. The observed sample.. in the system is a stochastic incomplete data, and the missing data.. exhibits a multilevel hierarchical data structure. The em-algorithm to find ML-estimates for parameters in the stochastic system is derived. Applications of the em-algorithm are exhibited in the two DTMC models, to find ML-estimates of the system parameters. Numerical results are given for influenza epidemics in the state of Georgia (GA), USA.
Due to the increase in volatile power generation facilities, the need for flexible modeling options of an energy network is growing. One approach consists of a cellular architecture whose hierarchy levels are less pro...
详细信息
Due to the increase in volatile power generation facilities, the need for flexible modeling options of an energy network is growing. One approach consists of a cellular architecture whose hierarchy levels are less pronounced. Such an architecture is provided by the Loop Circle Arc theory (LoCA theory). Each cell consists of essentially uniform basic building blocks, such as a storage unit, an energy converter, and a source and load, as well as an interface to the next cell. Based on this theory, a model of N households connected to a Circle is created. In order to report the demand of the connected households to the next cell, the Arc, via the interface, it is necessary to know the summed power values. Since the households generally represent stochastic processes, the densities associated with the households are estimated under the assumption of measured consumption values over a 24-hour period. Using the em-algorithm, mixed distribution densities are estimated based on normal distribution densities for each household and superimposed accordingly. In this way, in addition to the expected total power consumption, a variance can be given at the same time. This allows not only an estimation of the energy to be made available at certain times. It is also possible to simplify the network, since the N households can be approximated by the time evolution of the expected overall power consumption values. (C) 2021 Published by Elsevier Ltd.
Modern infocommunication networks are developing very intensively, the structure of networks and the nature of applications are changing, which leads to the need for new approaches to research the parameters of their ...
详细信息
ISBN:
(纸本)9781728135649
Modern infocommunication networks are developing very intensively, the structure of networks and the nature of applications are changing, which leads to the need for new approaches to research the parameters of their functioning. The paper proposes the solution to the problem of approximating an arbitrary probability density in the G/G/1 system using hyperexponential distributions. Moreover, to determine the parameters of the probability density of the hyperexponential distribution, it is proposed to use the emalgorithm. An example of approximation by a system of type H-2/D-2/1 of a real implementation of IP network traffic is given.
The using of Gabor function and texture features for pattern recognition in satellite images, principal component analysis (pca) and em-algorithm are discussed in this paper. Based on the presented in this article res...
详细信息
ISBN:
(纸本)9781509067428
The using of Gabor function and texture features for pattern recognition in satellite images, principal component analysis (pca) and em-algorithm are discussed in this paper. Based on the presented in this article results of the experiment, the algorithm based on em-algorithm and Gabor wavelets demonstrated high quality of the result and low time characteristics. The novelty of the method under consideration is the use of the Gabor wavelet to isolate texture features in conjunction with the use of the em-algorithm for image segmentation and the pca algorithm to reduce the dimensionality of the feature space of the extracted images.
Current status data appear in many biomedical studies when we only know if an event of interest occurs before or after a specific time point. In this paper, we develop statistical inference for the estimation of param...
详细信息
Current status data appear in many biomedical studies when we only know if an event of interest occurs before or after a specific time point. In this paper, we develop statistical inference for the estimation of parameters from current status data under the Lindley lifetime distribution, which is seen to work better than the exponential distribution in some lifetime contexts. We first develop an emalgorithm for Maximum Likelihood (ML) estimation and derive the asymptotic confidence intervals for model parameters. Then, we address the problem of model misspecification and define a new family of robust divergence-based estimators as a robust alternative to ML. Finally, we illustrate these methods through a simulation study as well as a numerical example.
In this paper, we introduce a bivariate distribution on R(+)xN arising from a single underlying Markov jump process. The marginal distributions are phase-type and discrete phase-type distributed, respectively, which a...
详细信息
In this paper, we introduce a bivariate distribution on R(+)xN arising from a single underlying Markov jump process. The marginal distributions are phase-type and discrete phase-type distributed, respectively, which allow for flexible behavior for modeling purposes. We show that the distribution is dense in the class of distributions on R(+)xN and derive some of its main properties, all explicit in terms of matrix calculus. Furthermore, we develop an effective emalgorithm for the statistical estimation of the distribution parameters. In the last part of the paper, we apply our methodology to an insurance dataset, where we model the number of claims and the mean claim sizes of policyholders, which is seen to perform favorably. An additional consequence of the latter analysis is that the total loss size in the entire portfolio is captured substantially better than with independent phase-type models.
The gamma-Poisson and beta-binomial mixture distributions are used for analyzing count-valued data, and the estimation of the hyper-parameters including the shape and/or scale parameters is important in the empirical ...
详细信息
The gamma-Poisson and beta-binomial mixture distributions are used for analyzing count-valued data, and the estimation of the hyper-parameters including the shape and/or scale parameters is important in the empirical Bayes inference. The maximum likelihood method requires the nested loops for solving the non-linear equations at each step of iteration in the emalgorithm. To avoid the extra loops, we derive the closed-form updating procedures at each step of iteration by using the score-adjusted method. The performance is compared by simulation with the maximum likelihood estimators.
The pace in the development and adoption of the new technologies for bigdata analytics has changed dramatically over the last several decades, and the amount of data being digitally ingested and stored is expanding ex...
详细信息
The pace in the development and adoption of the new technologies for bigdata analytics has changed dramatically over the last several decades, and the amount of data being digitally ingested and stored is expanding exponentially and rapidly. These data include structured, semi-structured and unstructured, and come in different sizes and formats. To utilize these vast resources, the knowledge and the skills needed to manage and to convert it into information is crucial. In this paper, firstly, the commonly used technologies, platforms, computational tools and the techniques currently in use for the ingesting, processing, storing and analyzing bigdata are reviewed. Secondly, those technologies are utilized to predict internet congestion by employing the bivariate mixture transition distribution (BMTD), expectation-maximization (em) algorithm and the autoregressive integrated moving average (ARIMA) models. BMTD models are very effective in capturing non-Gaussian and nonlinear features, such as bursts of activity and outliers, in a single unified model class. These models do not assume equally spaced, as well as independence, which are the key weaknesses of some other available time series and marked point processes models. Both the Weibull BMTD and the ARIMA models are very effective time series predictive models, but the comparison of their predictive performances is not yet addressed in the statistics and the machine learning literature.
暂无评论