Fruchterman Reingold method compares differently to agglomerative clustering
Interrelating Fruchterman-Reingold Graph Visualization and Agglomerative Clustering
Machine Learning
Summary
When looking at data, two ways to understand it are by drawing graphs or grouping data points together, called clustering. This paper looks at how a popular graph drawing method, Fruchterman-Reingold, relates to four common types of clustering. The authors found that these clustering methods usually agree with each other but have only moderate similarity with the original data. The graph drawing method kept more resemblance to the original data but was less like the clustering results. This suggests these methods see data patterns in related but distinct ways.
What this means in practice
- •For data analysts: Use graph visualization alongside clustering to gain complementary views of high-dimensional data patterns.
- •For machine learning engineers: Incorporate insights from graph visualization to better understand clustering outcomes in pattern recognition tasks.
Authors
Alexandre Benatti, Luciano da F. Costa
Abstract
Graph visualization methods and agglomerative clustering have been frequently considered in data analysis and pattern recognition. Because these approaches are interrelated and complementary, it is of particular interest to investigate their associations. In this work, we study the possible relationship between the Fruchterman-Reingold graph visualization method and four types of agglomerative clustering adopting single- and complete-linkage, average, and Ward's linkage criteria. Three types of datasets have been considered in 2 and 10 dimensions, as well as the PCA projection of the latter to two dimensions. The results obtained suggest that the relationship between the methods considered did not vary much for the three types of data mentioned above. At the same time, the agglomerative methods tended to yield results that are mostly similar to each other, while presenting moderate similarity with the original data. The Fruchterman-Reingold visualization resulted similar to the original data, but exhibited relatively smaller similarity to the agglomerative methods.