☆ 4.4 Article

Dirichlet ARMA models for compositional time series

JOURNAL OF MULTIVARIATE ANALYSIS (2017)

Journal

JOURNAL OF MULTIVARIATE ANALYSIS

Volume 158, Issue -, Pages 31-46

Publisher

ELSEVIER INC

DOI: 10.1016/j.jmva.2017.03.006

Keywords

Compositional data; Dirichlet distribution; Gaussian pseudo-likelihood; Multivariate time series; UK gross final expenditure series; Vector ARMA model

Categories

Statistics & Probability

Funding

National Natural Science Foundation of China [71371160]
Program for Changjiang Youth Scholar
Program for New Century Excellent Talents in University [NCET-13-0509]
National Science Foundation [DMS-1513409, DMS-1209085]

Ask authors/readers for more resources

Protocol

Community support

Reagent

Community support

Abstract

A compositional time series is a multivariate time series in which the observation vector at each time point is a set of proportions that sum to 1. Traditionally, such time series are modeled by taking a log-ratio transformation of the observations and then modeling them with a Gaussian vector autoregressive moving average (ARMA) model. In this paper, a new class of models is proposed by assuming that the proportions follow a time-varying Dirichlet distribution, and that the corresponding time-varying parameters, after a proper transformation, assume an ARMA-type of dynamic structure. The new model is referred to as the Dirichlet autoregressive moving average (DARMA) model. Under this model, after a proper transformation, the original data follow a vector ARMA model with a martingale difference sequence as its noise series. Two specific transformations are studied under the DARMA framework. Estimation procedures are developed and their numerical properties are investigated. Simulation studies and real examples are presented to demonstrate the properties of the proposed models, and comparisons are made with the existing modeling approaches. (C) 2017 Elsevier Inc. All rights reserved.

Authors

I am an author on this paper

Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.4

Not enough ratings

Secondary Ratings

Novelty

-

Significance

-

Scientific rigor

-

Rate this paper

Recommended

Article Economics

Compositional tourists' expenditure: Modeling through Dirichlet regression

Juan Gabriel Brida, Bibiana Lanzilotta, Leonardo Moreno

Summary: This study examines the relative expenditure patterns of tourists based on budget allocation and the influence of covariates. It introduces a model to characterize and compare different types of tourists in terms of their budget distribution. The empirical analysis of data from the Inbound Tourism survey in Uruguay reveals that the expenditure patterns in accommodation, food, and other items depend on factors such as destination, season, nationality, and type of accommodation. The inferential analysis also provides insights into different typologies of tourists from a microeconomic perspective.

TOURISM ECONOMICS (2023)

Add to Collection

Article Computer Science, Artificial Intelligence

A classification framework for multivariate compositional data with Dirichlet feature embedding

Jie Gu, Bin Cui, Shan Lu

Summary: This paper proposes an effective framework for multivariate compositional data classification, utilizing Dirichlet feature embedding to remove data constraint, obtain high-quality training data, and reduce dimensionality, followed by employing support vector machine to build the classification model. Results from simulation study and real-world dataset demonstrate the proposed method achieves good performance.

KNOWLEDGE-BASED SYSTEMS (2021)

Add to Collection

Article Computer Science, Artificial Intelligence

Change point detection for compositional multivariate data

Prabuchandran K. J., Nitin Singh, Pankaj Dayama, Ashutosh Agarwal, Vinayaka Pandit

Summary: Change point detection algorithms have a wide range of applications and the paper proposes a parametric approach for change point detection in compositional data. The method performs significantly better on compositional data and is competitive on general data compared to state-of-the-art implementations.

APPLIED INTELLIGENCE (2022)

Add to Collection

Article Computer Science, Theory & Methods

High-dimensional structure learning of sparse vector autoregressive models using fractional marginal pseudo-likelihood

Kimmo Suotsalo, Yingying Xu, Jukka Corander, Johan Pensar

Summary: In this paper, an approximate Bayesian approach combining fractional marginal likelihood and pseudo-likelihood is adopted to learn vector autoregressive models. A novel method, PLVAR, is proposed, which is faster and produces more accurate estimates than state-of-the-art methods based on penalized regression. The consistency of the PLVAR estimator is proven, and the method demonstrates attractive performance on both simulated and real-world data.

STATISTICS AND COMPUTING (2021)

Add to Collection

Article Statistics & Probability

The Integrated Nested Laplace Approximation for Fitting Dirichlet Regression Models

Joaquin Martinez-Minaya, Finn Lindgren, Antonio Lopez-Quilez, Daniel Simpson, David Conesa

Summary: This article introduces a Laplace approximation method for Bayesian inference in Dirichlet regression models, which allows analyzing skewed and heteroscedastic variables on a simplex without data transformation. The article provides theoretical foundations, implementation details, and introduces the dirinla package in R-language for Dirichlet regression. Simulation studies validate the proposed method, and a real data case-study demonstrates its application.

JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS (2023)

Add to Collection

Article Statistics & Probability

Observation-driven models for discrete-valued time series

Mirko Armillotta, Alessandra Luati, Monia Lupparelli

Summary: This paper develops inference for a general class of first order observation-driven models for discrete-valued processes. Stochastic properties are derived under easy-to-check conditions, and consistency and asymptotic normality of quasi-maximum likelihood estimators are established.

ELECTRONIC JOURNAL OF STATISTICS (2022)

Add to Collection

Article Biology

Estimation of prediction error in time series

Alexander Aue, Prabir Burman

Summary: This article presents a method for accurately estimating prediction errors in time series, including both univariate and multivariate stationary time series. By starting from an approximation to the bias-variance decomposition of the prediction error, several estimates are derived for various popular time series models. Simulation experiments show that the proposed estimators perform well in finite samples and can be used for model selection and prediction.

BIOMETRIKA (2023)

Add to Collection

Article Anthropology

Bayesian nonparametric mixtures of Exponential Random Graph Models for ensembles of networks

Sa Ren, Xue Wang, Peng Liu, Jian Zhang

Summary: In this manuscript, a method called Dirichlet Process Mixture of Exponential Random Graph Models (DPM-ERGMs) is proposed to model ensembles of networks. The method divides the ensemble into different clusters and models each cluster using a separate Exponential Random Graph Model (ERGM). The number of clusters can be determined automatically and adaptively with the data provided. A Metropolis-within-slice sampling algorithm is developed for full Bayesian inference of DPM-ERGMs.

SOCIAL NETWORKS (2023)

Add to Collection

Article Computer Science, Theory & Methods

Quantile hidden semi-Markov models for multivariate time series

Luca Merlo, Antonello Maruotti, Lea Petrella, Antonio Punzo

Summary: This paper develops a new method for jointly estimating multiple quantiles of multivariate time series. The method considers the correlation structure among outcomes and models the unobserved serial heterogeneity using a hidden semi-Markov chain. Inference is carried out using an efficient algorithm that avoids parametric assumptions about the states' sojourn distributions.

STATISTICS AND COMPUTING (2022)

Add to Collection

Article Statistics & Probability

A SELF-NORMALIZED APPROACH TO SEQUENTIAL CHANGE-POINT DETECTION FOR TIME SERIES

Ngai Hang Chan, Wai Leong Ng, Chun Yip Yau

Summary: A self-normalization sequential change-point detection method is proposed for time series analysis, aiming to address issues with traditional tests such as sensitivity to bandwidth parameters and size distortion. The null asymptotic and consistency of the proposed method are established under general regularity conditions, and its effectiveness is demonstrated through simulation experiments and application to railway-bearing temperature data.

STATISTICA SINICA (2021)

Add to Collection

Article Statistics & Probability

Robust bootstrap prediction intervals for univariate and multivariate autoregressive time series models

Ufuk Beyaztas, Han Lin Shang

Summary: This paper proposes a robust bootstrap algorithm for constructing prediction intervals and forecast regions in autoregressive time series models. The algorithm is based on weighted likelihood estimates and weighted residuals, and its finite sample properties are examined through Monte Carlo studies and empirical data examples.

JOURNAL OF APPLIED STATISTICS (2022)

Add to Collection

Article Statistics & Probability

Generalized Gaussian quasi-maximum likelihood estimation for most common time series

Yakoub Boularouk, Jean-Marc Bardet

Summary: This study proposes a Generalized Gaussian Quasi-Maximum Likelihood Estimator for estimating the parameter shape of the generalized gaussian noise in the class of causal time series. Monte Carlo experiments confirm the accuracy of the proposed estimator.

COMMUNICATIONS IN STATISTICS-THEORY AND METHODS (2022)

Add to Collection

Article Health Care Sciences & Services

A pseudo-likelihood approach for multivariate meta-analysis of test accuracy studies with multiple thresholds

Annamaria Guolo, Duc-Khanh To

Summary: A new method using pseudo-likelihood for inference on common diagnostic measures is proposed in this study, which assumes working independence between sensitivities and specificities at different thresholds in the same study, overcoming some of the issues in multivariate meta-analysis. Simulation studies demonstrate the satisfactory performance of the method under different scenarios, significantly improving results compared to the multivariate normal counterpart.

STATISTICAL METHODS IN MEDICAL RESEARCH (2021)

Add to Collection

Article Economics

Modeling Multivariate Time Series With Copula-Linked Univariate D-Vines

Zifeng Zhao, Peng Shi, Zhengjun Zhang

Summary: This article introduces a novel multivariate time series model named CuDvine, which models both temporal and cross-sectional dependence simultaneously. By linking multiple uDvines, CuDvine provides a flexible model that can handle complex time series dependency issues.

JOURNAL OF BUSINESS & ECONOMIC STATISTICS (2022)

Add to Collection

Article Economics

A new BISARMA time series model for forecasting mortality using weather and particulate matter data

Victor Leiva, Helton Saulo, Rubens Souza, Robert G. Aykroyd, Roberto Vila

Summary: The paper introduces a new class of time series models, called BS autoregressive moving average (BISARMA) models, based on the Birnbaum-Saunders (BS) distribution, which allows modeling of positive and asymmetric data with an autoregressive structure. Through a thorough study of the theoretical properties and practical issues of the proposed methodology, performance evaluation using Monte Carlo simulations, and analysis of real-world data, the potential of the method is demonstrated. The numerical results show the excellent performance of the BISARMA model, indicating that the BS distribution is a good modeling choice for time series data with positive support and asymmetric distribution.

JOURNAL OF FORECASTING (2021)

Add to Collection

No Data Available

Article Statistics & Probability

On moments of truncated multivariate normal/independent distributions

Tsung- Lin, Wan-Lun Wang

Summary: This paper derives explicit expressions for the moments of truncated multivariate normal/independent distributions with supports confined within a hyper-rectangle. A Monte Carlo experiment is conducted to validate the proposed formulae for five selected members of the distributions.

JOURNAL OF MULTIVARIATE ANALYSIS (2024)

Add to Collection

Article Statistics & Probability

Testing homogeneity in high dimensional data through random projections

Tao Qiu, Qintong Zhang, Yuanyuan Fang, Wangli Xu

Summary: This article introduces a method for testing the homogeneity of two random vectors. The method involves selecting two subspaces and projecting them onto one-dimensional spaces, using the Cramer-von Mises distance to construct the test statistic. The performance is enhanced by repeating this procedure and the effectiveness is demonstrated through numerical simulations.

JOURNAL OF MULTIVARIATE ANALYSIS (2024)

Add to Collection

Article Statistics & Probability

Matrix-valued isotropic covariance functions with local extrema

Alfredo Alegria, Xavier Emery

Summary: This study contributes to covariance modeling by proposing new parametric families of isotropic matrix-valued functions that exhibit non-monotonic behaviors, such as hole effects and cross-dimples. The benefit of these models is demonstrated on a bivariate dataset of airborne particulate matter concentrations.

JOURNAL OF MULTIVARIATE ANALYSIS (2024)

Add to Collection

Article Statistics & Probability

Asymptotic properties of hierarchical clustering in high-dimensional settings

Kento Egashira, Kazuyoshi Yata, Makoto Aoshima

Summary: This study investigates the asymptotic properties of hierarchical clustering in different settings, including high-dimensional, low-sample-size scenarios. The results show that hierarchical clustering exhibits good asymptotic properties under practical settings for high-dimensional data. The study also extends the analysis to consider scenarios where both the dimension and sample size approach infinity, and generalizes the concept of populations in multiclass HDLSS settings.

JOURNAL OF MULTIVARIATE ANALYSIS (2024)

Add to Collection

Article Statistics & Probability

Quantile-based MANOVA: A new tool for inferring multivariate data in factorial designs

Marlene Baumeister, Marc Ditzhaus, Markus Pauly

Summary: This paper introduces a more robust multivariate analysis method by using general quantiles, particularly the median, instead of the traditional mean, and applies and validates this method on various factorial designs. The effectiveness of this method is demonstrated through theoretical and simulation studies on small and moderate sample sizes.

JOURNAL OF MULTIVARIATE ANALYSIS (2024)

Add to Collection

Article Statistics & Probability

Stochastic representations and probabilistic characteristics of multivariate skew-elliptical distributions

Chuancun Yin, Narayanaswamy Balakrishnan

Summary: The family of multivariate skew-normal distributions has interesting properties, which also hold for a general class of skew-elliptical distributions.

JOURNAL OF MULTIVARIATE ANALYSIS (2024)

Add to Collection

Article Statistics & Probability

On testing the equality of latent roots of scatter matrices under ellipticity

Gaspard Bernard, Thomas Verdebout

Summary: In this paper, we address the problem of testing the relationship between the eigenvalues of a scatter matrix in an elliptical distribution. Using the Le Cam asymptotic theory, we show that the non-specification of nuisance parameters has an asymptotic cost for testing the relationship. We also propose a distribution-free signed-rank test for this problem.

JOURNAL OF MULTIVARIATE ANALYSIS (2024)

Add to Collection

© Peeref 2019-2024. All rights reserved.