Relationship between Diversity and Perfomance of Multiple Classifiers for Decision Support

The paper presents the investigation and implementation of the relationship between diversity and the performance of multiple classifiers on classification accuracy. The study is critical as to build classifiers that are strong and can generalize better. The parameters of the neural network within the committee were varied to induce diversity; hence structural diversity is the focus for this study. The hidden nodes and the activation function are the parameters that were varied. The diversity measures that were adopted from ecology such as Shannon and Simpson were used to quantify diversity. Genetic algorithm is used to find the optimal ensemble by using the accuracy as the cost function. The results observed shows that there is a relationship between structural diversity and accuracy. It is observed that the classification accuracy of an ensemble increases as the diversity increases. There was an increase of 3%-6% in the classification accuracy.

• Publication year

  2008

• Venue

  arXiv.org

• Publication date

  2008-10-21

• Fields of study

  Computer Science

• Identifiers
  arXiv 0810.3865
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• The Effect of Structural Diversity of an Ensemble of Classifiers on Classification Accuracy

  2008cited by this paper
• Exploiting Diversity in Ensembles: Improving the Performance on Unbalanced Datasets

  2007cited by this paper
• Bayesian Training of Neural Networks Using Genetic Programming

  2006cited by this paper
• Pattern Recognition and Machine Learning

  2006influential reference
• Evolutionary Ensemble Creation and Thinning

  2006influential reference
• Autoencoder networks for HIV classification

  2006cited by this paper
• Optimal ensemble construction via meta-evolutionary ensembles

  2006cited by this paper
• Diversity creation methods: a survey and categorisation

  2004influential reference
• The Conceptualization and Measurement of Diversity

  2003cited by this paper
• Measures of Diversity in Classifier Ensembles and Their Relationship with the Ensemble Accuracy

  2003influential reference
• Relationships between combination methods and measures of diversity in combining classifiers

  2002cited by this paper
• Ten measures of diversity in classifier ensembles: limits for two classifiers

  2001cited by this paper
• Bayesian neural networks for classification: how useful is the evidence framework?

  1999cited by this paper
• Popular Ensemble Methods: An Empirical Study

  1999cited by this paper
• On Combining Classifiers

  1998influential reference

• Regularization and Learning an Ensemble of RNNs by Decorrelating Representations

  2017cites this paper
• Financial credit risk assessment: a recent review

  2015cites this paper
• A Consensus Approach for Combining Multiple Classifiers in Cost-Sensitive Bankruptcy Prediction

  2013cites this paper