Standard audiogram classification from loudness scaling data using unsupervised, supervised, and explainable machine learning techniques