ACTGNN: Assessment of Clustering Tendency with Synthetically-Trained Graph Neural Networks

2025-01-30

AI Summary

Existing clustering tendency assessment methods (e.g., Hopkins statistic, VAT) exhibit poor robustness and heavy reliance on subjective visual interpretation when applied to large-scale, high-dimensional, noisy data with weak cluster structure. Method: We propose the first graph neural network (GNN) framework for unsupervised clustering tendency assessment: node features are encoded via locality-sensitive hashing (LSH), edge features are constructed using multiple similarity metrics (e.g., RBF kernel), and the model is trained self-supervisedly on synthetic data—requiring no ground-truth labels to capture intrinsic clustering propensity. Contribution/Results: Extensive experiments demonstrate that our method significantly outperforms mainstream baselines on both synthetic and diverse real-world datasets. It reliably detects weak cluster structures even under low signal-to-noise ratios and high dimensionality, achieving superior robustness, scalability, and label-free adaptability.

Technology Category

• Machine Learning: Clustering
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Normalization, clustering, classification, and summarization of Web text
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Determining clustering tendency in datasets is a fundamental but challenging task, especially in noisy or high-dimensional settings where traditional methods, such as the Hopkins Statistic and Visual Assessment of Tendency (VAT), often struggle to produce reliable results. In this paper, we propose ACTGNN, a graph-based framework designed to assess clustering tendency by leveraging graph representations of data. Node features are constructed using Locality-Sensitive Hashing (LSH), which captures local neighborhood information, while edge features incorporate multiple similarity metrics, such as the Radial Basis Function (RBF) kernel, to model pairwise relationships. A Graph Neural Network (GNN) is trained exclusively on synthetic datasets, enabling robust learning of clustering structures under controlled conditions. Extensive experiments demonstrate that ACTGNN significantly outperforms baseline methods on both synthetic and real-world datasets, exhibiting superior performance in detecting faint clustering structures, even in high-dimensional or noisy data. Our results highlight the generalizability and effectiveness of the proposed approach, making it a promising tool for robust clustering tendency assessment.

Problem

• High-Dimensional Data
• Clustering Tendency
• Noise Handling

Innovation

• Graph Neural Networks
• Radial Basis Function Kernel
• Locality Sensitive Hashing