Deep Learning Method Based for Breast Cancer Classification

The most prevalent cancer in women worldwide and one of the main factors in cancer-related mortality is breast cancer. Extensive research efforts have been dedicated to early detection, diagnosis, and treatment of breast cancer to reduce mortality rates. This research aims to achieve accurate diagnosis of breast cancer and classify breast cancer using deep learning method. The study proposes deep learning techniques with Adam's optimization and two hidden layers for breast cancer classification. Addressing challenges such as data instability and overfitting during deep learning training, the research focuses on updating network weights. The experiments examine two hidden layers and varying learning rates to enhance classification accuracy. The datasets utilized in the experiments include the WBCD dataset (Original), the WDBC dataset (Diagnostics), and the Coimbra dataset. Additionally, the proposed scheme's accuracy is compared against existing benchmarks for breast cancer detection. The experimental findings show that the suggested scheme outperforms other benchmarks, achieving an impressive 96% accuracy in breast cancer classification.