AI Paper Atlas

Research map / Learning theory and algorithms / Matrix and tensor methods

Tensor Graphical Model: Non-Convex Optimization and Statistical Inference

Xiang Lyu, Will Wei Sun, Zhaoran Wang, Han Liu, Jian Yu Yang, Guang Cheng

TPAMI 2019 · Learning theory and algorithms › Matrix and tensor methods · approach: High-dimensional estimation theory · 36 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use tensor-structured estimation (Tlasso) instead of vectorized GlassoTlasso · Simulations s1-s3Kronecker precision estimation error and TPR: 0.343 -> 1Improved
replace P-MLE iterative updates with one-step TlassoTlasso · Scenario s3computation time: P-MLE takes about one minute vs about three secondsImproved
replace direct Glasso with TlassoTlasso · Scenario s3computation time: more than half an hour -> about three secondsImproved
use Tlasso instead of P-MLE for precision matrix estimationTlasso · Scenarios s1-s3Frobenius and max norm errorImproved
increase sample size from n=50 to n=80Tlasso · Scenarios s1 to s2estimation errorImproved
increase one dimension m3 from 10 to 20Tlasso · Scenarios s1 to s3Frobenius vs max norm errorMixed
use FDP control procedure instead of P-MLE for support recoveryTlasso with FDP control · Simulations s1-s3TNR: TNR 0.893 -> 0.935 (Sim1 s1)Improved
use data-driven instead of oracle covariance in inferenceTlasso inference · Simulations s1-s3FDP and power: FDP 7.8 vs 6.7 and power 100 vs 100 (Sim1 s1, 5%)No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy