ClusBench: The Clustering Benchmark Data Resource You've All Been Waiting For (?)
Researchers have introduced ClusBench, a new benchmark data resource designed to improve the evaluation of clustering methods. This resource comprises nearly 3000 synthetic datasets generated from over 200 real-world datasets, retaining the complexity of original data while allowing for larger scale benchmarking. The synthetic datasets and an accompanying R package are publicly available for download. AI
IMPACT Provides a more robust and scalable evaluation framework for clustering algorithms, potentially leading to improved model development.