Addressing the Fundamental Tension of PCGML with Discriminative Learning

Isaac Karth, Adam M. Smith

Procedural content generation via machine learning (PCGML) is typically framed as the task of fitting a generative model to full-scale examples of a desired content distribution. T…