Scalable unsupervised feature selection via weight stability

Xudong Zhang, Renato Cordeiro de Amorim

Unsupervised feature selection is critical for improving clustering performance in high-dimensional data, where irrelevant features can obscure meaningful structure. In this work,…