An expectation–maximization algorithm for the matrix normal distribution with an application in remote sensing

2018 ◽  
Vol 167 ◽  
pp. 31-48 ◽  
Author(s):  
Hunter Glanz ◽  
Luis Carvalho
Keyword(s):  
Remote Sensing ◽  
Normal Distribution ◽  
Expectation Maximization ◽  
Expectation Maximization Algorithm ◽  
The Matrix ◽  
Matrix Normal ◽  
Matrix Normal Distribution

scholarly journals On n/p-Asymptotic Distribution Of Vector Of Weighted Traces Of Powers Of Wishart Matrices

2018 ◽  
Vol 33 ◽  
pp. 24-40 ◽  
Author(s):  
Jolanta Pielaszkiewicz ◽  
Dietrich Von Rosen ◽  
Martin Singull
Keyword(s):  
Normal Distribution ◽  
Covariance Matrix ◽  
Asymptotic Distribution ◽  
Joint Distribution ◽  
Recursive Formula ◽  
Multivariate Normal ◽  
Wishart Matrices ◽  
The Matrix ◽  
Matrix Normal ◽  
Matrix Normal Distribution

The joint distribution of standardized traces of $\frac{1}{n}XX'$ and of $\Big(\frac{1}{n}XX'\Big)^2$, where the matrix $X:p\times n$ follows a matrix normal distribution is proved asymptotically to be multivariate normal under condition $\frac{{n}}{p}\overset{n,p\rightarrow\infty}{\rightarrow}c>0$. Proof relies on calculations of asymptotic moments and cumulants obtained using a recursive formula derived in Pielaszkiewicz et al. (2015). The covariance matrix of the underlying vector is explicitely given as a function of $n$ and $p$.

scholarly journals A Bayesian Multiple-Trait and Multiple-Environment Model Using the Matrix Normal Distribution

2018 ◽  
Author(s):  
Osval A. Montesinos-López ◽  
Abelardo Montesinos-López ◽  
José Cricelio Montesinos-López ◽  
José Crossa ◽  
Francisco Javier Luna-Vázquez ◽  
...  
Keyword(s):  
Normal Distribution ◽  
Multiple Trait ◽  
Environment Model ◽  
The Matrix ◽  
Matrix Normal ◽  
Matrix Normal Distribution

scholarly journals Dynamic Brain Connectivity Alternation Detection via Matrix-variate Differential Network Model

2018 ◽  
Author(s):  
Jiadong Ji ◽  
Yong He ◽  
Lei Xie
Keyword(s):  
Normal Distribution ◽  
Partial Correlation ◽  
Graphical Model ◽  
Brain Connectivity ◽  
Connectivity Analysis ◽  
Differential Network ◽  
The Matrix ◽  
Matrix Normal ◽  
Matrix Normal Distribution ◽  
The Brain

AbstractMotivationNowadays brain connectivity analysis has attracted tremendous attention and has been at the foreground of neuroscience research. Brain functional connectivity reveals the synchronization of brain systems through correlations in neurophysiological measures of brain activity. Growing evidence now suggests that the brain connectivity network experiences alternations with the presence of numerous neurological disorders, thus differential brain network analysis may provides new insights into disease pathologies. For the matrix-valued data in brain connectivity analysis, existing graphical model estimation methods assume a vector normal distribution that in essence requires the columns of the matrix data to be independent. It is obviously not true, they have limited applications. Among the few solutions on graphical model estimation under a matrix normal distribution, none of them tackle the estimation of differential graphs across different populations. This motivates us to consider the differential network for matrix-variate data to detect the brain connectivity alternation.ResultsThe primary interest is to detect spatial locations where the connectivity, in terms of the spatial partial correlation, differ across the two groups. To detect the brain connectivity alternation, we innovatively propose a Matrix-Variate Differential Network (MVDN) model. MVDN assumes that the matrix-variate data follows a matrix-normal distribution. We exploit the D-trace loss function and a Lasso-type penalty to directly estimate the spatial differential partial correlation matrix where the temporal information is fully excavated. We propose an ADMM algorithm for the Lasso penalized D-trace loss optimization problem. We investigate theoretical properties of the estimator. We show that under mild and regular conditions, the proposed method can identify all differential edges accurately with probability tending to 1 in high-dimensional setting where dimensions of matrix-valued data p, q and sample size n are all allowed to go to infinity. Simulation studies demonstrate that MVDN provides more accurate differential network estimation than that achieved by other state-of-the-art methods. We apply MVDN to Electroencephalography (EEG) dataset, which consists of 77 alcoholic individuals and 45 controls. The hub genes and differential interaction patterns identified are consistent with existing experimental [email protected] informationSupplementary data are available online.

Hybrid fuzzy interface model of sports rehabilitation activities

2021 ◽  
pp. 1-10
Author(s):  
Wu Shoujiang
Keyword(s):  
Evaluation System ◽  
Clustering Algorithm ◽  
Mixed Model ◽  
Expectation Maximization Algorithm ◽  
Rehabilitation Training ◽  
Development Platform ◽  
Longitudinal Tracking ◽  
Database Structure ◽  
The Matrix ◽  
Matrix Normal

At present, the relevant test data and training indicators of athletes during rehabilitation training lack screening and analysis, so it is impossible to establish a long-term longitudinal tracking research system and evaluation system. In order to improve the practical effect of sports rehabilitation activities, this paper successively introduces the matrix normal mixed model and the fuzzy clustering algorithm based on the K-L information entropy regularization and the matrix normal mixed model. Moreover, this paper uses the expectation maximization algorithm to estimate the parameters of the model, discusses the framework, key technologies and core services of the development platform, and conducts certain research on the related technologies of the three-tier architecture. At the same time, according to the actual needs of sports rehabilitation training, this paper designs the functions required for exercise detection and prescription formulation. In addition, this paper analyzes and designs the database structure involved in each subsystem. Finally, this paper designs experiments to verify the performance of the model constructed in this paper. The research results show that the performance of the model constructed in this paper meets the expectations of model construction, so it can be applied to practice.

scholarly journals Diagnostic Analytics for an Autoregressive Model under the Skew-Normal Distribution

Mathematics ◽  
2020 ◽  
Vol 8 (5) ◽  
pp. 693 ◽  
Author(s):  
Yonghui Liu ◽  
Guohua Mao ◽  
Víctor Leiva ◽  
Shuangzhe Liu ◽  
Alejandra Tapia
Keyword(s):  
Time Series ◽  
Normal Distribution ◽  
Expectation Maximization ◽  
Autoregressive Model ◽  
Expectation Maximization Algorithm ◽  
Model Fit ◽  
Simulation Studies ◽  
Skew Normal Distribution ◽  
Financial Return ◽  
Skew Normal

Autoregressive models have played an important role in time series. In this paper, an autoregressive model based on the skew-normal distribution is considered. The estimation of its parameters is carried out by using the expectation–maximization algorithm, whereas the diagnostic analytics are conducted by means of the local influence method. Normal curvatures for the model under four perturbation schemes are established. Simulation studies are conducted to evaluate the performance of the proposed procedure. In addition, an empirical example involving weekly financial return data are analyzed using the procedure with the proposed diagnostic analytics, which has improved the model fit.

Sign in / Sign up

Export Citation Format

Share Document