Semisupervised Learning for a Hybrid Generative/Discriminative Classifier based on the Maximum Entropy Principle

Authors:
Akinori Fujino;Naonori Ueda;Kazumi Saito
Affiliations:
-;-;-
Venue:
IEEE Transactions on Pattern Analysis and Machine Intelligence
Year:
2008

Citing 0
Cited 11

Computational intelligence in gait research: a perspective on current applications and future challenges

IEEE Transactions on Information Technology in Biomedicine - Special section on computational intelligence in medical systems
Tag confidence measure for semi-automatically updating named entity recognition

NEWS '09 Proceedings of the 2009 Named Entities Workshop: Shared Task on Transliteration
Help-Training for semi-supervised support vector machines

Pattern Recognition
Locally discriminative topic modeling

Pattern Recognition
Efficient semi-supervised learning on locally informative multiple graphs

Pattern Recognition
Logistic label propagation

Pattern Recognition Letters
Topic evolution prediction of user generated contents considering enterprise generated contents

Proceedings of the First ACM International Workshop on Hot Topics on Interdisciplinary Social Networks Research
Combining information theoretic kernels with generative embeddings for classification

Neurocomputing
Bidirectional semi-supervised learning with graphs

ECML PKDD'12 Proceedings of the 2012 European conference on Machine Learning and Knowledge Discovery in Databases - Volume Part II
A second order cone programming approach for semi-supervised learning

Pattern Recognition
On the characterization of noise filters for self-training semi-supervised in nearest neighbor classification

Neurocomputing

Quantified Score

Hi-index	0.14

Visualization

Abstract

This paper presents a method for designing semi-supervised classifiers trained on labeled and unlabeled samples. We focus on probabilistic semi-supervised classifier design for multi-class and singlelabeled classification problems, and propose a hybrid approach that takes advantage of generative and discriminative approaches. In our approach, we first consider a generative model trained by using labeled samples and introduce a bias correction model, where these models belong to the same model family, but have different parameters. Then, we construct a hybrid classifier by combining these models based on the maximum entropy principle. To enable us to apply our hybrid approach to text classification problems, we employed naive Bayes models as the generative and bias correction models. Our experimental results for four text data sets confirmed that the generalization ability of our hybrid classifier was much improved by using a large number of unlabeled samples for training when there were too few labeled samples to obtain good performance. We also confirmed that our hybrid approach significantly outperformed generative and discriminative approaches when the performance of the generative and discriminative approaches was comparable. Moreover, we examined the performance of our hybrid classifier when the labeled and unlabeled data distributions were different.