Title :
Instance based random forest with rotated feature space
Author :
Le Zhang ; Ye Ren ; Suganthan, P.
Author_Institution :
Sch. of Electr. & Electron. Eng., Nanyang Technol. Univ., Singapore, Singapore
Abstract :
Random Forest is a competitive ensemble method in the field of machine learning with several advantages such as efficiency, robustness, generalization, ease of implementation, etc. This study attempts to increase the diversity among the pairwise individuals in the forest. On the other hand, we propose an instance based method to select several superior trees to perform the voting. The proposed method is evaluated on 28 datasets from the UCI Repository.
Keywords :
learning (artificial intelligence); pattern classification; UCI repository; competitive ensemble method; instance based method; instance based random forest; machine learning; rotated feature space; Accuracy; Bagging; Boosting; Principal component analysis; Testing; Training; Vegetation;
Conference_Titel :
Computational Intelligence and Ensemble Learning (CIEL), 2013 IEEE Symposium on
Conference_Location :
Singapore
DOI :
10.1109/CIEL.2013.6613137
