Integration of Stochastic Models by Minimizing α-Divergence

DOI10.1162/neco.2007.19.10.2780zbMath1138.62002WikidataQ80836076 ScholiaQ80836076MaRDI QIDQ5441322

Shun-ichi Amari

Publication date: 11 February 2008

Published in: Neural Computation (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1162/neco.2007.19.10.2780

zbMATH Keywords

mixture of experts; mixtures of distributions; \(\alpha\)-divergence; \(\alpha\)-mixtures

Mathematics Subject Classification ID

68T05: Learning and adaptive systems in artificial intelligence

91E40: Memory and learning in psychology

62B10: Statistical aspects of information-theoretic topics

Related Items

CONFORMAL GEOMETRY OF ESCORT PROBABILITY AND ITS APPLICATIONS, Multiclass Alpha Integration of Scores from Multiple Classifiers, Parameter Estimation for α-GMM Based on Maximum Likelihood Criterion, Parameter Learning for Alpha Integration, Fusion of Scores in a Detection Context Based on Alpha Integration, A Note on Divergences, Non-linear canonical correlation analysis using alpha-beta divergence, The information geometry of Bregman divergences and some applications in multi-expert reasoning, On clustering histograms with \(k\)-means by using mixed \(\alpha\)-divergences, A risk profile for information fusion algorithms, Saddlepoint condition on a predictor to reconfirm the need for the assumption of a prior distribution, Unsupervised weight parameter estimation method for ensemble learning, Quantum Hellinger distances revisited, An Estimation of Generalized Bradley-Terry Models Based on the em Algorithm

Cites Work

• Second order efficiency of minimum contrast estimators in a curved exponential family
• I-divergence geometry of probability distributions and minimization problems
• Possible generalization of Boltzmann-Gibbs statistics.
• Illumination-invariance of plateau's midgray
• Training Products of Experts by Minimizing Contrastive Divergence
• On asymptotic properties of predictive distributions
• On the local geometry of mixture models
• The α-EM algorithm: surrogate likelihood maximization using α-logarithmic information measures
• Stochastic Reasoning, Free Energy, and Information Geometry
• Divergence Function, Duality, and Convex Analysis
• Information Geometry of U-Boost and Bregman Divergence
• A Generalized Bayes Rule for Prediction
• Means of Positive Numbers and Matrices
• GENERALIZATION OF THE MEAN-FIELD METHOD FOR POWER-LAW DISTRIBUTIONS
• A Measure of Asymptotic Efficiency for Tests of a Hypothesis Based on the sum of Observations