A FOCUSED INFORMATION CRITERION FOR GRAPHICAL MODELS IN FMRI CONNECTIVITY WITH HIGH-DIMENSIONAL DATA

被引：9

作者：

Pircalabelu, Eugen ^{[1
,2
]}

Claeskens, Gerda ^{[1
,2
]}

Jahfari, Sara ^{[3
]}

Waldorp, Lourens J. ^{[4
]}

机构：

[1] Katholieke Univ Leuven, Fac Econ & Business, ORSTAT, Naamsestr 69, B-3000 Leuven, Belgium

[2] Katholieke Univ Leuven, Fac Econ & Business, Leuven Stat Res Ctr, B-3000 Leuven, Belgium

[3] Vrije Univ Amsterdam, Fac Psychol & Educ, Dept Cognit Psychol, Boechorststr 1, NL-1081 BT Amsterdam, Netherlands

[4] Univ Amsterdam, Fac Social & Behav Sci, Dept Psychol Methods, NL-1018 Amsterdam, Netherlands

来源：

ANNALS OF APPLIED STATISTICS | 2015年 / 9卷 / 04期

关键词：

fMRI connectivity; focused information criterion; model selection; Gaussian; graphical model; penalization; high-dimensional data; NONCONCAVE PENALIZED LIKELIHOOD; STATE FUNCTIONAL CONNECTIVITY; VARIABLE SELECTION; ADAPTIVE LASSO; SMALL-WORLD; BRAIN; NETWORKS; CORTEX; REGRESSION; ROBUST;

D O I：

10.1214/15-AOAS882

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

Connectivity in the brain is the most promising approach to explain human behavior. Here we develop a focused information criterion for graphical models to determine brain connectivity tailored to specific research questions. All efforts are concentrated on high-dimensional settings where the number of nodes in the graph is larger than the number of samples. The graphical models may include autoregressive times series components, they can relate graphs from different subjects or pool data via random effects. The proposed method selects a graph with a small estimated mean squared error for a user-specified focus. The performance of the proposed method is assessed on simulated data sets and on a resting state functional magnetic resonance imaging (fMRI) data set where often the number of nodes in the estimated graph is equal to or larger than the number of samples.

引用

页码：2179 / 2214

页数：36

共 50 条

[1] A High-dimensional Focused Information Criterion
Gueuning, Thomas
Claeskens, Gerda
[J]. SCANDINAVIAN JOURNAL OF STATISTICS, 2018, 45 (01) : 34 - 61
[2] A focused information criterion for graphical models
Eugen Pircalabelu
Gerda Claeskens
Lourens Waldorp
[J]. Statistics and Computing, 2015, 25 : 1071 - 1092
[3] A focused information criterion for graphical models
Pircalabelu, Eugen
Claeskens, Gerda
Waldorp, Lourens
[J]. STATISTICS AND COMPUTING, 2015, 25 (06) : 1071 - 1092
[4] ESTIMATION OF HIGH-DIMENSIONAL CONNECTIVITY IN FMRI DATA VIA SUBSPACE AUTOREGRESSIVE MODELS
Ting, Chee-Ming
Seghouane, Abd-Krim
Salleh, Sh-Hussain
[J]. 2016 IEEE STATISTICAL SIGNAL PROCESSING WORKSHOP (SSP), 2016,
[5] The robust desparsified lasso and the focused information criterion for high-dimensional generalized linear models
Pandhare, S. C.
Ramanathan, T. V.
[J]. STATISTICS, 2023, 57 (01) : 1 - 25
[6] Constructing Graphical Models via the Focused Information Criterion
Claeskens, Gerda
Pircalabelu, Eugen
Waldorp, Lourens
[J]. MODELING AND STOCHASTIC LEARNING FOR FORECASTING IN HIGH DIMENSIONS, 2015, 217 : 55 - 78
[7] High-dimensional undirected graphical models for arbitrary mixed data
Goebler, Konstantin
Drton, Mathias
Mukherjee, Sach
Miloschewski, Anne
[J]. ELECTRONIC JOURNAL OF STATISTICS, 2024, 18 (01): : 2339 - 2404
[8] High-Dimensional Mixed Graphical Models
Cheng, Jie
Li, Tianxi
Levina, Elizaveta
Zhu, Ji
[J]. JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, 2017, 26 (02) : 367 - 378
[9] High-dimensional Gaussian graphical models on network-linked data
Li, Tianxi
Qian, Cheng
Levina, Elizaveta
Zhu, Ji
[J]. 1600, Microtome Publishing (21):
[10] Inference and Verification of Probabilistic Graphical Models from High-Dimensional Data
Ma, Yinjiao
Damazyn, Kevin
Klinger, Jakob
Gong, Haijun
[J]. DATA INTEGRATION IN THE LIFE SCIENCES, DILS 2015, 2015, 9162 : 223 - 239

← 1 2 3 4 5 →