We present a new parallel DBSCAN algorithm for Spark on the Google Cloud Platform. Statistical analysis is applied to determine DBSCAN's optimal parameters to enhance clustering performance. for scalability Cost-based, ...