An Enhanced Probabilistic Neural Network Approach Applied to Text Classification

Authors:
Patrick Marques Ciarelli;Elias Oliveira
Affiliations:
Universidade Federal do Espírito Santo, Vitória, Brazil 29075-910;Universidade Federal do Espírito Santo, Vitória, Brazil 29075-910
Venue:
CIARP '09 Proceedings of the 14th Iberoamerican Conference on Pattern Recognition: Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications
Year:
2009

Citing 8
Cited 0

Probabilistic neural networks

Neural Networks
BoosTexter: A Boosting-based Systemfor Text Categorization

Machine Learning - Special issue on information retrieval
Machine learning in automated text categorization

ACM Computing Surveys (CSUR)
Modern Information Retrieval

Modern Information Retrieval
Pattern Classification (2nd Edition)

Pattern Classification (2nd Edition)
ML-KNN: A lazy learning approach to multi-label learning

Pattern Recognition
A Comparison between a KNN Based Approach and a PNN Algorithm for a Multi-label Classification Problem

ISDA '08 Proceedings of the 2008 Eighth International Conference on Intelligent Systems Design and Applications - Volume 02
Learning multi-label alternating decision trees from texts and data

MLDM'03 Proceedings of the 3rd international conference on Machine learning and data mining in pattern recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

Text classification is still a quite difficult problem to be dealt with both by the academia and by the industrial areas. On the top of that, the importance of aggregating a set of related amount of text documents is steadily growing in importance these days. The presence of multi-labeled texts and great quantity of classes turn this problem even more challenging. In this article we present an enhanced version of Probabilistic Neural Network using centroids to tackle the multi-label classification problem. We carried out some experiments comparing our proposed classifier against the other well known classifiers in the literature which were specially designed to treat this type of problem. By the achieved results, we observed that our novel approach were superior to the other classifiers and faster than the Probabilistic Neural Network without the use of centroids.