Simulated Iterative Classification A New Learning Procedure for Graph Labeling

Authors:
Francis Maes;Stéphane Peters;Ludovic Denoyer;Patrick Gallinari
Affiliations:
LIP6 - University Pierre et Marie Curie, Paris, France;LIP6 - University Pierre et Marie Curie, Paris, France;LIP6 - University Pierre et Marie Curie, Paris, France;LIP6 - University Pierre et Marie Curie, Paris, France
Venue:
ECML PKDD '09 Proceedings of the European Conference on Machine Learning and Knowledge Discovery in Databases: Part II
Year:
2009

Citing 13
Cited 3

On the foundations of relaxation labeling processes

Readings in computer vision: issues, problems, principles, and paradigms
A maximum entropy approach to natural language processing

Computational Linguistics
Why collective inference improves relational classification

Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining
Learning structured prediction models: a large margin approach

ICML '05 Proceedings of the 22nd international conference on Machine learning
Ranking on graph data

ICML '06 Proceedings of the 23rd international conference on Machine learning
Linear prediction models with graph regularization for web-page categorization

Proceedings of the 12th ACM SIGKDD international conference on Knowledge discovery and data mining
Manifold Regularization: A Geometric Framework for Learning from Labeled and Unlabeled Examples

The Journal of Machine Learning Research
Classification in Networked Data: A Toolkit and a Univariate Case Study

The Journal of Machine Learning Research
Know your neighbors: web spam detection using the web topology

SIGIR '07 Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval
Stacked dependency networks for layout document structuring

Proceedings of the 2008 ACM symposium on Applied computing
Stacked sequential learning

IJCAI'05 Proceedings of the 19th international joint conference on Artificial intelligence
Regularization on discrete spaces

PR'05 Proceedings of the 27th DAGM conference on Pattern Recognition
Iterative decoding of compound codes by probability propagation in graphical models

IEEE Journal on Selected Areas in Communications

Classification and annotation in social corpora using multiple relations

Proceedings of the 20th ACM international conference on Information and knowledge management
Transforming graph data for statistical relational learning

Journal of Artificial Intelligence Research
Learning latent representations of nodes for classifying in heterogeneous social networks

Proceedings of the 7th ACM international conference on Web search and data mining

Quantified Score

Hi-index	0.00

Visualization

Abstract

Collective classification refers to the classification of interlinked and relational objects described as nodes in a graph. The Iterative Classification Algorithm (ICA) is a simple, efficient and widely used method to solve this problem. It is representative of a family of methods for which inference proceeds as an iterative process: at each step, nodes of the graph are classified according to the current predicted labels of their neighbors. We show that learning in this class of models suffers from a training bias. We propose a new family of methods, called Simulated ICA, which helps reducing this training bias by simulating inference during learning. Several variants of the method are introduced. They are both simple, efficient and scale well. Experiments performed on a series of 7 datasets show that the proposed methods outperform representative state-of-the-art algorithms while keeping a low complexity.