.png){kind=icon}
.png){kind=icon}
Return
Basis precision matrix estimation via column-wise inverse operator
.png){kind=icon}
PRE
AI
DOI：10.1016/j.spl.2026.110685.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This paper proposes an improved column-wise inverse operator method for precision matrix estimation of high-dimensional compositional data. We construct the estimator by using the truncated sample log-ratio covariance matrix and provide a probability bound on Gq(s(n,p) , M- n,M-p ) under 4+epsilon ( epsilon > 0) moment condition. Simulations and a human gut microbiome dataset analysis confirm that the method works well.
Keywords:
High-dimensional compositional data
Basis precision matrix estimation
Improved SCIO method
Low moment condition
Journal
S
IF:
0.7
Papers:
121
Citations:
0