An adaptive semi-supervised stochastic forest algorithm based on data similarity is proposed. The coding and mapping of labeled and unlabeled data are carried out by using random forest. The pseudo-marking of unlabeled data is selected. Then the randomized forest is trained to improve the classification performance. The experimental results show that the proposed algorithm can effectively use the unmarked data to improve the classification accuracy.
关键词
Keywords
references
Leistner C, Saffari A, Santner J, et al. Semi-Supervised Random Forests[C]// IEEE, International Conference on Computer Vision. IEEE, 2010: 506-513.
Liu X, Song M, Tao D, et al. Semi-supervised node splitting for random forest construction[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2013: 492-499.
Blum A, Mitchell T. Combining labeled and unlabeled data with co-training[C]// Eleventh Conference on Computational Learning Theory. ACM, 1998: 92-100.
Goldman S, Zhou Y. Enhancing supervised learning with unlabeled data[C]//ICML. 2000: 327-334.
Sch?lkopf B, Platt J, Hofmann T. Fast Discriminative Visual Codebooks using Randomized Clustering Forests[C]// International Conference on Neural Information Processing Systems. MIT Press, 2006: 985-992.
Deng C, Guo M Z. A new co-training-style random forest for computer aided diagnosis[J]. Journal of Intelligent Information Systems, 2011, 36(3):253-281. DOI: 253
Breiman L. Random forests[J].Machine learning, 2001, 5(1):5-32. DOI: 5
Geurts P, Ernst D, Wehenkel L. Extremely randomized trees[J]. Machine Learning, 2006, 63(1):3-42. DOI: 3