Parallel Visual Assessment of Cluster Tendency on GPU

Abstract

Determining the number of clusters in a data set is a critical issue in cluster analysis. The Visual Assessment of (cluster) Tendency (VAT) algorithm is an effective tool for investigating cluster tendency, which produces an intuitive image of matrix as the representation of complex data sets. However, VAT can be computationally expensive for large data sets due to its \( O\left( {N^{2} } \right) \) time complexity. In this paper, we propose an efficient parallel scheme to accelerate the original VAT using NVIDIA GPU and CUDA architecture. We show that, on a range of data sets, the GPU-based VAT features good scalability and can achieve significant speedups compared to the original algorithm.