Fig. 1From: Measuring the model risk-adjusted performance of machine learning algorithms in credit default predictionThe dilemma between prediction gains and algorithmic complexityBack to article page