KAUST Research Workshop on Optimization and Big Data
École des Ponts - ParisTech
Guillaume Obozinski is a researcher in the Computer Science department at Ecole des Ponts ParisTech since 2013. He was previously part of the SIERRA INRIA project-team, in the Computer Science department of the Ecole Normale Supérieure, Paris, France. A former student of the Ecole Normale Supérieure in Cachan, he earned his PhD in 2009 from the Statistics department of the University of California at Berkeley. His research interests include machine learning, statistics, optimization, sparsity and their applications.
I'll present an efficient dual augmented Lagrangian formulation to learn conditional random field (CRF) models. The algorithm, which can be interpreted as an inexact gradient method on the multiplier, does not require to perform exact inference iteratively, requires only a fixed number of stochastic clique-wise updates at each epoch to obtain a sufficiently good estimate of the gradient w.r.t. the Lagrange multipliers. We prove that the proposed algorithm enjoys global linear convergence for both the primal and the dual objective. Our experiments show that the proposed algorithm outperforms state-of-the-art baselines in terms of speed of convergence. (Joint work with Shell Xu Hu)