On Structured Prediction Theory with Calibrated Convex Surrogate Losses.

Anton Osokin Francis R. Bach Simon Lacoste-Julien

Published in: NIPS (2017)

Keyphrases

structured prediction
convex optimization
maximum margin
latent variables
conditional random fields
markov networks
efficient learning
high order
convex hull
information retrieval
probability distribution
random variables
approximate inference