Summary: This Australian Research Council funded discovery project has developed new ensemble classifier methods. The methods are based on clustering of input data space, optimising base-classifiers and fusing them. The developed ensemble methods have been rigorously tested on numerous benchmark and real-world datasets and achieved classification accuracy better than the existing state-of-the-art ensemble methods. The research has shown the impact of clustering input space, types of base classifiers, numbers of base classifiers and diversity of base classifiers on ensemble accuracy in developing optimised ensemble classifiers.
Presentation and Demo: 10 June 2020, 1 pm (Brisbane time), Zoom meeting: https://cqu.zoom.us/j/96487227843, Media Release (Click Here)
Code and Data: The code and data for developed ensemble classifier methods are available on GitHub. You can download Code and Data (Click Here). We encourage you to run the code with different data including your own data and let us know any new results you get. Any questions or feedback can be sent to Professor Brijesh Verma by email (b.verma@cqu.edu.au or b.verma.qld@gmail.com).
Publications: This research project has produced 16 journal and conference papers including IEEE Transactions on Evolutionary Computation, IEEE Transaction of Neural Networks and Learning Systems, Neurocomputing, Pattern Recognition and IEEE International Joint Conference on Neural Networks. It has also produced 1 PhD thesis.
Research Training: This research project has provided research training to 2 postdoc research fellows and 1 PhD student.
Selected Papers
1. S. Fletcher, B. Verma and M. Zhang. A Non-Specialized Ensemble Classifier using Multi-Objective Optimization, Neurocomputing, 2020. Available Online
2. Z. Jan and B. Verma, Multi-Cluster Class Balanced Ensemble, IEEE Transactions on Neural Networks and Learning Systems, 2020. Available Online
3. Z. Jan and B. Verma, "Evolutionary Classifier and Cluster Selection Approach for Ensemble Classification," ACM Transactions on Knowledge Discovery in Data, vol. 14, no. 1, pp. 1-8, 2019. Available Online
4. M. Asafuddoula, B. Verma and M. Zhang, A Divide-and-Conquer Based Ensemble Classifier Learning using Many-Objective Optimization, IEEE Transactions on Evolutionary Computation, vol. 22, no. 5, pp. 762-777, 2018. Available Online
5. S. Fletcher, B. Verma, Z. Jan and M. Zhang, The Optimized Selection of Base-Classifiers for Ensemble Classification using a Multi-Objective Genetic Algorithm, International Joint Conference on Neural Networks (IJCNN), pp. 1-8, 2018. Available Online
6. M. Asafuddoula, B. Verma and M. Zhang, An Incremental Ensemble Classifier Learning by Means of a Rule-Based Accuracy and Diversity Comparison, IEEE International Joint Conference on Neural Networks (IJCNN), pp. 1924-1931, 2017. Available Online
7. S. Fletcher and B. Verma, Removing Bias from Diverse Data Clusters for Ensemble Classification. 24th International Conference on Neural Information Processing (ICONIP), pp. 140–149, 2017, Springer. Available Online