AI Paper Atlas

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

The Search Problem in Mixture Models

Avik Ray, Joe Neeman, Sujay Sanghavi, Sanjay Shakkottai

JMLR 2018 · Learning theory and algorithms › Matrix and tensor methods · approach: High-dimensional estimation theory · 9 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 side-information Whitening search instead of seeded k-meansGMM image segmentation · BSDS500 testmedian NMI: 0.13 -> 0.17Improved
use Whitening instead of seeded k-meansGMM image segmentation · BSDS500 trainmedian NMI: 0.15 -> 0.15No clear effect
use Whitening instead of seeded k-meansGMM image segmentation · BSDS500 valmedian NMI: 0.09 -> 0.11No clear effect
use Whitening instead of seeded k-meansGMM image segmentation · BSDS500 testmedian runtime (s): 81.5 -> 6.7Improved
use Whitening instead of tensor power methodtopic model on NY Times (k=10) · NY Timesruntime (s): 908 -> 188Improved
use Whitening instead of NMF and SS-NMFtopic modeling with labeled words · NY Times / YelpPMI score: 40 of 62 words; 35 of 54 wordsImproved
use Whitening instead of NMF and SS-NMFtopic modeling · NY Times / Yelpspeedup: more than 2XImproved
use Whitening/Cancellation instead of TPMGMM synthetic · k=10, d=500relative error gainImproved
initialize tensor power iteration with side information (Fast-TPM) (ablation)TPM · GMM syntheticerror gain over TPMNo clear effect
increase sample size n, proportion alpha and variance sigma (ablation)Whitening · GMM syntheticerror gain over TPMImproved
use Whitening on rare component (alpha_min .0037)Whitening · GMM k=10, d=500error gain over TPMImproved
use subspace search instead of k-meanssubspace clustering · Synthetic k=5, m=10, d=500speedup: over 4XImproved

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