A Survey of Transfer Learning in Breast Cancer Image Classification
Jinghong Xu, Xinyou Dong · 2020 IEEE 3rd International Conference of Safe Production and Informatization (IICSPI) · 2020
Worldwide, breast cancer is one of the leading cancer for female, especially in low and middle income countries. Breast cancer's classification in the early stages plays a pivotal role to improve survival rate. But the training dataset which should be large-scaled and well-annotated is hard to collect because of the limitation of data acquisition and annotation. Transfer learning can transfer the knowledge from target domain to source domain to improve model performance. Thus, the dataset needed for training in target domain can be degraded. For the feature mentioned above, it has become the hot topic for research in machine learning. This survey tries to systematize the current research of transfer learning techniques in the area of breast cancer image classification. And also, highlight several findings for the researchers in future.