The proposed work: New structure of CNN presented to detect breast cancer using AlexNet with AdaBoost. The proposed work is compared with previous studies and applies a new method based on the deep convolutional pretrained model named AlexNet. The AlexNet is a famous pretrained model used in various image classification problems. After applying AlexNet to extract high-level features from input images, the extracted features size will be 4096 in all AlexNet applications, which are wired to classifiers in traditional studies. However, in this study, we wire the extracted features by the AlexNet to the Energy Spectral Density (ESD), which is used as a feature extraction in various studies, and combine it with the CNN model.
I use MATLAB for programming. Please ensure that the MATLAB codes can be run completely and directly.
Read
Read Breast Data
Apply AlexNet
Energy Spectral Density (ESD)
Ensemble Learning
No
Train
Complete?
Yes
Test Model
Complete