Computerized breast cancer analysis system using three stage semi-supervised learning method

• The abundance of available unlabeled data can be used in this scheme.• We developed data weighing module and redesigned similarity function.• We designed and combined local features and global features together.• Our system has acceptable sensitivity even at a low false positive rate.

Source: Computer Methods and Programs in Biomedicine - July 7, 2016 Category: Bioinformatics Authors: Wenqing Sun, Tzu-Liang (Bill) Tseng, Jianying Zhang, Wei Qian Source Type: research