AI Paper Atlas

Research map / Learning theory and algorithms / Clustering

A Novel Normalized-Cut Solver With Nearest Neighbor Hierarchical Initialization

Feiping Nie, Jitao Lu, Danyang Wu, Rong Wang, Xuelong Li

TPAMI 2023 · Learning theory and algorithms › Clustering · approach: Spectral clustering algorithms · 34 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
solve N-Cut directly with Fast-CD instead of EVD+KMEVD+KM · MNIST10ACC: 0.6954 -> 0.7766Improved
solve N-Cut directly with Fast-CD instead of ISRISR · MNIST10ACC: 0.5768 -> 0.7766Improved
use Fast-CD instead of EVD+KMEVD+KM · 8 benchmark datasetsN-Cut objective: 8.5232 -> 8.6528 on MNIST10Improved
use Fast-CD instead of EVD+KMEVD+KM · SegmentNMI: 0.5088 -> 0.4875Hurt
use Fast-CD instead of EVD+KMEVD+KM · MSRA50NMI: 0.6933 -> 0.6734Hurt
use Fast-CD instead of ISRISR · STL-10execution time (s): 224.203 -> 0.574Improved
use Fast-CD instead of EVD+KMEVD+KM · MNIST10execution time (s): 2.702 -> 0.268Improved
use ISR instead of EVD+KM spectral discretizationEVD+KM · MNIST10ACC: 0.6954 -> 0.5768Hurt
refine EVD+KM labels with Fast-CD (ablation)EVD+KM · MNIST10NMI: 0.6506 -> 0.6784Improved
refine EVD+KM labels with Fast-CD (ablation)EVD+KM · GTdbNMI: 0.7021 -> 0.7004Hurt
refine EVD+KM labels with Fast-CD (ablation)EVD+KM · UMistN-Cut objective: 12.45199326 -> 12.59905529Improved
compare Fast-CD to SBMC/EBMC graph clusteringFast-CD · GTdbNMI: 0.7056 vs 0.7386Hurt

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