Comparing Compound and Ordinary Diversity measures Using Decision Trees.
Year: 2011
Type: dissertation
Abstract: An ensemble of classifiers succeeds in improving the accuracy of the whole when the\ncomponent classifiers are both diverse and accurate. Diversity is required to ensure that the\nclassifiers make uncorrelated errors. Theoretical and experimental approaches from previous\nresearch show very low correlation between ensemble accuracy and diversity measure.\nIntroducing Proposed Compound diversity functions by Albert Hung-Ren KO and Robert\nSabourin, (2009), by combining diversities and performances of individual classifiers exhibit\nstrong correlations between the diversities and accuracy. To be consistent with existing\narguments compound diversity of measures are evaluated and compared with traditional\ndiversity measures on different problems. Evaluating diversity of errors and comparison with\nmeasures are significant in this study. The results show that compound diversity measures\nare better than ordinary diversity measures. However, the results further explain evaluation of\ndiversity of errors on available data.
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W75789707
