A Lattice-Based Approach to Hierarchical Clustering

The paper presents an approach to hierarchical clustering based on the use of a least general generalization (lgg) operator to induce a lattice structure of clusters and a category utility objective function to evaluate the clustering quality. Theobjective function is integrated with a lattice-based distance measureinto a bottom-up control strategy for clustering. Experimentswith well-knowndatasets are discussed.

• Type: Conference
• Year: 2001
• Where: FLAIRS
• Authors: Zdravko Markov
• Keywords: Artificial Intelligence, Clustering Quality, FLAIRS 2001, Hierarchical Clustering, Utility Objective Function