conftrace_

2014 NeurIPS NeurIPS 2014

Distributed Parameter Estimation in Probabilistic Graphical Models

Abstract

This paper presents foundational theoretical results on distributed parameter estimation for undirected probabilistic graphical models. It introduces a general condition on composite likelihood decompositions of these models which guarantees the global consistency of distributed estimators, provided the local estimators are consistent.

📈 Trend Setter - Federated Learning

🐣 Hot Topic Early Bird - probabilistic graphical model

Authors

Yariv D Mizrahi, Misha Denil, Nando de Freitas

Topics

Artificial Intelligence > Bayesian & Probabilistic > Probabilistic Modeling Artificial Intelligence > Learning Paradigms > Federated Learning Machine Learning > Optimization & Theory > Distributed Learning Machine Learning > Optimization & Theory > Statistical Learning Machine Learning > Bayesian & Probabilistic > Probabilistic Modeling Machine Learning > Core Methods > Graphical Models

Keywords

probabilistic modeling parameter estimation distributed estimation markov random field graphical model probabilistic graphical model composite likelihood undirected graphical model distributed parameter estimation

More from NeurIPS 2014

All NeurIPS 2014 papers →

Estimation with Norm Regularization

A Multiplicative Model for Learning Distributed Text-Based Attribute Representations

Discovering Structure in High-Dimensional Data Through Correlation Explanation

Efficient Optimization for Average Precision SVM

From MAP to Marginals: Variational Inference in Bayesian Submodular Models

Analysis of Learning from Positive and Unlabeled Data