Benchmarking Parallel K-Means Cloud Type Clustering from Satellite Data

Abstract

The study of clouds, i.e., where they occur and what are their characteristics, plays a key role in the understanding of climate change. Clustering is a common machine learning technique used in atmospheric science to classify cloud types. Many parallelism techniques e.g., MPI, OpenMP and Spark, could achieve efficient and scalable clustering of large-scale satellite observation data. In order to understand their differences, this paper studies and compares three different approaches on parallel clustering of satellite observation data. Benchmarking experiments with k-means clustering are conducted with three parallelism techniques, namely OpenMP, OpenMP+MPI, and Spark, on a HPC cluster using up to 16 nodes.