Greedy part-wise learning of sum-product networks

Robert Peharz, Bernhard Geiger, Franz Pernkopf

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

35 Citaten (Scopus)


Sum-product networks allow to model complex variable interactions while still granting efficient inference. However, most learning algorithms proposed so far are explicitly or implicitly restricted to the image domain, either by assuming variable neighborhood or by assuming that dependent variables are related by their magnitudes over the training set. In this paper, we introduce a novel algorithm, learning the structure and parameters of sum-product networks in a greedy bottom-up manner. Our algorithm iteratively merges probabilistic models of small variable scope to larger and more complex models. These merges are guided by statistical dependence test, and parameters are learned using a maximum mutual information principle. In experiments our method competes well with the existing learning algorithms for sum-product networks on the task of reconstructing covered image regions, and outperforms these when neither neighborhood nor correlations by magnitude can be assumed.

Originele taal-2Engels
TitelMachine Learning and Knowledge Discovery in Databases European Conference, ECML PKDD 2013, Prague, Czech Republic, September 23-27, 2013
SubtitelProceedings Part 1
RedacteurenHendrik Blockeel, Kristian Kersting, Siegfried Nijssen, Filip Železný
ISBN van elektronische versie978-3-642-40991-2
StatusGepubliceerd - 2013
Extern gepubliceerdJa
Evenement2013 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2013) - Prague, Tsjechië
Duur: 23 sep 201327 sep 2013

Publicatie series

NaamLecture Notes in Computer Science book series
UitgeverijSpringer Link
ISSN van elektronische versie1973-2020
NaamLecture Notes in Artificial Intelligence book sub series


Congres2013 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2013)
Verkorte titelECML PKDD 2013


Duik in de onderzoeksthema's van 'Greedy part-wise learning of sum-product networks'. Samen vormen ze een unieke vingerafdruk.

Citeer dit