Lloyd's K-Means Algorithm Identified as Special Case of Frank-Wolfe Method

A new paper establishes a connection between Lloyd's K-Means Clustering Algorithm and the Frank-Wolfe (FW) algorithm, demonstrating that K-Means is a specific instance of FW. This research derives a non-asymptotic convergence rate for K-Means and proposes an FW variant to handle empty clusters, retaining the same convergence rate. The findings are illustrated with simulations on Gaussian mixtures and image segmentation data. AI

IMPACT Provides a theoretical framework that could lead to improved optimization methods for clustering tasks in machine learning.

RANK_REASON Academic paper detailing a novel theoretical connection between two algorithms. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv stat.ML →

• Frank-Wolfe algorithm
• Gaussian Mixture Models
• k-means clustering
• Lloyd's K-Means Clustering Algorithm

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →