An Upper Bound on the Global Optimum in Parameter Estimation
Abstract
Learning graphical model parameters from incomplete data is a non-convex optimization problem. Iterative algorithms, such as Expectation Maximization (EM), can be used to get a local optimum solution. However, little is known about the quality of the learned local optimum, compared to the unknown global optimum. We exploit variables that are always observed in the dataset to get an upper bound on the global optimum which can give insight into the quality of the parameters learned by estimation algorithms.
Cite
Text
Refaat and Darwiche. "An Upper Bound on the Global Optimum in Parameter Estimation." Conference on Uncertainty in Artificial Intelligence, 2015.Markdown
[Refaat and Darwiche. "An Upper Bound on the Global Optimum in Parameter Estimation." Conference on Uncertainty in Artificial Intelligence, 2015.](https://mlanthology.org/uai/2015/refaat2015uai-upper/)BibTeX
@inproceedings{refaat2015uai-upper,
title = {{An Upper Bound on the Global Optimum in Parameter Estimation}},
author = {Refaat, Khaled S. and Darwiche, Adnan},
booktitle = {Conference on Uncertainty in Artificial Intelligence},
year = {2015},
pages = {772-781},
url = {https://mlanthology.org/uai/2015/refaat2015uai-upper/}
}