Minority Class Oversampling for Tabular Data with Deep Generative Models