Building Diversified Multiple Trees for Classification in High Dimensional Noise Data

It is common that a trained classification model is applied to the operating data that is deviated from the training data because of noise. This paper demonstrate an ensemble classifier, Diversified Multiple Trees (DMT) is more robust to classify noised data than other widely used ensemble methods. DMT is tested on three real world biological data sets from different laboratories in comparison with four benchmark ensemble classifiers. Experimental results show that DMT is significantly more accurate than other benchmark ensemble classifiers on noised test data. We also discussed a limitation of DMT and its possible variations.