5 papers
An interpretable Good--Turing restart criterion for k-means++
Renato Cordeiro de Amorim
The k-means++ algorithm is commonly restarted multiple times to avoid poor local optima, yet the number of restarts is almost always chosen arbitrarily and applied uniformly regard…
Counterfactuals for Feature-Weighted Clustering
Richard J. Fawley, Renato Cordeiro de Amorim
Counterfactual explanations provide local, interpretable insight by identifying changes to an input that would alter its assigned outcome. Although well established in supervised l…
Cluster-Adaptive Feature Extraction and its Theoretical Foundation with Minkowski Weighted k-Means
Renato Cordeiro de Amorim, Vladimir Makarenkov
The Minkowski weighted -means (-means) algorithm extends classical -means by incorporating feature weights and a Minkowski distance. We first show that the -means o…
Shapley-Inspired Feature Weighting in -means with No Additional Hyperparameters
Richard J. Fawley, Renato Cordeiro de Amorim
Clustering algorithms often assume all features contribute equally to the data structure, an assumption that usually fails in high-dimensional or noisy settings. Feature weighting…
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,…