On-the-Go Adaptability in the New Ant Colony Decision Forest Approach
In this article we present a new and effective adaptive ant colony algorithm that we employ to construct a decision forest (aACDF). The aim of this proposition is to create an adaptive meta-ensemble based on data sets created during the algorithm runtime. This on-the-go approach allows to construct classifiers in which wrongly classified objects obtain a greater probability of being chosen for the pseudo-samples in subsequent iterations. Every pseudo-sample is created on the basis of training data. Our results confirm the standpoint that this new adaptive ACDF slightly reduces the accuracy of classification as well as builds considerably smaller decision trees.
KeywordsAnt Colony Decision Forest Boosting Ant Colony Optimization Decision Forest ACDT
Unable to display preview. Download preview PDF.
- 6.Corne, D., Dorigo, M., Glover, F.: New Ideas in Optimization. Mc Graw–Hill, Cambridge (1999)Google Scholar
- 7.Dorigo, M., Maniezzo, V., Colorni, A.: The ant system: an autocatalytic optimization process. Tech. Rep. 91-016, Department of Electronics, Politecnico di Milano, Italy (1996)Google Scholar
- 9.Dorigo, M., Birattari, M., Blum, C., Clerc, M., Stützle, T., Winfield, A.F.T. (eds.): ANTS 2008. LNCS, vol. 5217. Springer, Heidelberg (2008)Google Scholar
- 10.Freund, Y., Schapire, R.E.: Experiments with a new boosting algorithm. In: International Conference on Machine Learning, pp. 148–156 (1996)Google Scholar
- 12.Kearns, M.: Thoughts on hypothesis boosting, project for Ron Rivest’s machine learning course at MIT (1988)Google Scholar
- 13.Rokach, L., Maimon, O.: Data Mining With Decision Trees: Theory and Applications. World Scientific Publishing (2008)Google Scholar
- 15.Schapire, R.E.: The strength of weak learnability. Machine Learning 5, 197–227 (1990)Google Scholar