Result: Hybrid TLBO-PSO Algorithm Optimized Deep Learning Techniques for Analyzing Mammograms

Breast cancer which is the commonest malignant tumor in women, not only is a threat to life but also affects the mental and physical health of women. One of the most important tools in diagnosing breast cancer is Mammography. As mammogram images are complex, doctors find it difficult to identify the attributes of breast cancer clearly. The classification algorithm which is being used to study breast cancer at present is deep learning. So, this work proposes a Residual Network (ResNet) 34 and Convolution Neural Network (CNN) 18 model for benign as well as malignant mammographic images’ proper as well as precise classification. Teaching-Learning Based Optimization algorithm (TLBO) with Particle Swarm Optimization (PSO) (TLBO-PSO), a fundamental deep learning approach, has been used in this study. This approach’s key goal is for optimization of the outcome of the solution vectors on the CNN as well as the ResNet so as to enhance precision or recognition. The accuracy of this model not only helps in better performance and enhanced accuracy of malignant and benign classification of mammogram images but also proves the robustness and generalization of the model.