On Structured Prediction Theory with Calibrated Convex Surrogate Losses

Open in new window