Clustering Approaches for Mixed-Type Data: A Comparative Study

2025-11-24

AI Summary

Existing clustering methods for mixed-type data (comprising both continuous and categorical variables) lack systematic performance evaluation, particularly regarding their robustness under varying data characteristics. Method: We conduct a comprehensive simulation study across multiple factors—variable interaction strength, cluster overlap, sample size, and proportion of continuous variables—and benchmark six prominent algorithms: k-prototypes, KAMILA, LCM, PDQ, convex k-means, and mixture models based on Bayesian networks. Adjusted Rand Index (ARI) serves as the unified evaluation metric. Contribution/Results: KAMILA, LCM, and k-prototypes achieve the highest overall performance. Strong variable dependencies substantially degrade clustering accuracy across all methods, exposing fundamental limitations. Cluster overlap and the proportion of continuous variables emerge as the most influential factors affecting performance. All implementations are publicly available in R, ensuring full reproducibility.

Technology Category

Machine Learning: ClusteringConstraint Satisfaction and Optimization: Mixed Discrete/Continuous OptimizationData Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

Web Mining and Content Analysis: Robustness and generalizability of Web mining methodsGraph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphsSemantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics

Abstract

Clustering is widely used in unsupervised learning to find homogeneous groups of observations within a dataset. However, clustering mixed-type data remains a challenge, as few existing approaches are suited for this task. This study presents the state-of-the-art of these approaches and compares them using various simulation models. The compared methods include the distance-based approaches k-prototypes, PDQ, and convex k-means, and the probabilistic methods KAy-means for MIxed LArge data (KAMILA), the mixture of Bayesian networks (MBNs), and latent class model (LCM). The aim is to provide insights into the behavior of different methods across a wide range of scenarios by varying some experimental factors such as the number of clusters, cluster overlap, sample size, dimension, proportion of continuous variables in the dataset, and clusters' distribution. The degree of cluster overlap and the proportion of continuous variables in the dataset and the sample size have a significant impact on the observed performances. When strong interactions exist between variables alongside an explicit dependence on cluster membership, none of the evaluated methods demonstrated satisfactory performance. In our experiments KAMILA, LCM, and k-prototypes exhibited the best performance, with respect to the adjusted rand index (ARI). All the methods are available in R.

Problem

• Comparing clustering methods for mixed-type data across diverse scenarios
• Evaluating performance impact of cluster overlap and variable proportions
• Identifying limitations when variables interact with cluster dependencies

Innovation

• Compares clustering methods for mixed-type data
• Evaluates distance-based and probabilistic approaches
• Identifies top performers using adjusted rand index