Learning graphical model parameters with approximate marginal inference

Justin Domke*

*Corresponding author for this work

    Research output: Contribution to journalArticlepeer-review

    127 Citations (Scopus)

    Abstract

    Likelihood-based learning of graphical models faces challenges of computational complexity and robustness to model misspecification. This paper studies methods that fit parameters directly to maximize a measure of the accuracy of predicted marginals, taking into account both model and inference approximations at training time. Experiments on imaging problems suggest marginalization-based learning performs better than likelihood-based approximations on difficult problems where the model being fit is approximate in nature.

    Original languageEnglish
    Article number6420841
    Pages (from-to)2454-2467
    Number of pages14
    JournalIEEE Transactions on Pattern Analysis and Machine Intelligence
    Volume35
    Issue number10
    DOIs
    Publication statusPublished - 2013

    Fingerprint

    Dive into the research topics of 'Learning graphical model parameters with approximate marginal inference'. Together they form a unique fingerprint.

    Cite this