Reproduce “K-GBS3FCM -- KNN Graph-Based Safe Semi-Supervised Fuzzy C-Means” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Gabriel Santos, Rita Julia, Marcelo Nascimento
K-GBS3FCM, developed by researchers at the Federal University of Uberlândia, introduces a safe semi-supervised fuzzy c-means clustering algorithm that leverages KNN graph structures and dynamic parameter adjustments to achieve superior accuracy and robustness against mislabeled data, even at mislabeling ratios up to 30%.
Key insights
- Dynamic neighborhood size is determined based on local data density rather than fixed parameters, allowing the algorithm to adapt to varying data distributions.
- A simplified and computationally efficient method is used for calculating safety degrees, reflecting label reliability based on consistency with neighboring unlabeled points.
- The integration of KNN graph structures into the objective function and adaptive weighting mechanisms enhances robustness against mislabeling and better captures spatial data relationships.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.