Learning efficient random maximum a-posteriori predictors with non-decomposable loss functions
Name
Jaakkola_Learning efficient.pdf
Size
354.39 KB
Format
Adobe PDF
Checksum (MD5)
020ac103e4afc50cd8d569ec20f76634
Author(s)
Hazan, Tamir
Maji, Subhransu
Keshet, Joseph
Jaakkola, Tommi S.
Date Issued
2013
Journal
Advances in Neural Information Processing Systems (NIPS)
Publisher
Neural Information Processing Systems
Citation
Hazan, Tamir, Subhransu Maji, Joseph Keshet, and Tommi Jaakkola. "Learning efficient random maximum a-posteriori predictors with non-decomposable loss functions." Advances in Neural Information Processing Systems (NIPS 2013).
Version
Final published version
Abstract
In this work we develop efficient methods for learning random MAP predictors for structured label problems. In particular, we construct posterior distributions over perturbations that can be adjusted via stochastic gradient methods. We show that every smooth posterior distribution would suffice to define a smooth PAC-Bayesian risk bound suitable for gradient methods. In addition, we relate the posterior distributions to computational properties of the MAP predictors. We suggest multiplicative posteriors to learn super-modular potential functions that accompany specialized MAP predictors such as graph-cuts. We also describe label-augmented posterior models that can use efficient MAP approximations, such as those arising from linear program relaxations.
MIT Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.
Persistent DSpace Link
DOI of Published Version
http://papers.nips.cc/paper/5066-learning-efficient-random-maximum-a-posteriori-predictors-with-non-decomposable-loss-functions
