The Effect of Correlation on the Accuracy of Meta-Learning Approach

Authors:
Li-ying Yang
Affiliations:
Xiýan Jiaotong University
Venue:
ICALT '05 Proceedings of the Fifth IEEE International Conference on Advanced Learning Technologies
Year:
2005

Citing 4
Cited 0

Original Contribution: Stacked generalization

Neural Networks
Cascade Generalization

Machine Learning
Generating classifier outputs of fixed accuracy and diversity

Pattern Recognition Letters
Measures of Diversity in Classifier Ensembles and Their Relationship with the Ensemble Accuracy

Machine Learning

Quantified Score

Hi-index	0.00

Visualization

Abstract

Meta-learning is an efficient approach in the field of machine learning, which involves multiple classifiers. In this paper, a meta-learning framework consisting of stacking meta-learning and cascade meta-learning was proposed firstly. Then the algorithm for generating simulated datasets was presented. Finally, based on the classifier simulator, datasets with variable correlation were obtained and used to evaluate the classification performance of meta-learning. Experimental results show that negative correlation measured by Q statistic benefits meta-learning approach.