How many trees in a random forest?
MLDM'12 Proceedings of the 8th international conference on Machine Learning and Data Mining in Pattern Recognition
Root attribute behavior within a random forest
IDEAL'12 Proceedings of the 13th international conference on Intelligent Data Engineering and Automated Learning
Hi-index | 0.00 |
This paper introduces an efficient keyword based medical image retrieval method using image classification and confidence assigning of each keyword. To classify images, we first extract wavelet-based CSLBP (WCS-LBP) descriptors from local parts of the images and then we apply the extracted feature vector to decision trees to construct random forests, which are an ensemble of random decision trees. For semantic annotation based on classification results, we propose the confidence assigning method to keywords according to probabilities of random forests with predefined body relation graph (BRG). After keyword annotation with different confidence, we proved that our keyword based image retrieval method showed more efficient retrieval results compared to equal confidence method.