Breast cancer forecast and diagnosis using machine learning approaches: A comparative analysis
Ranjeet Kumar Dubey, Rajesh Kumar Singh, Shashank Srivastav, Anshu Kumar Dwivedi · 2023
In recent times, breast cancer has been the disease that affects most women worldwide. Because people aren’t aware of the first signs of cancer, the death rate from breast cancer keeps rising. It is already possible to identify breast cancer using various tools and environments with real-time algorithms. All training and classification fields have seen a rapid rise in terms of application of machine learning. In modern computer programming methods for categorizing breast cancer, the Deep Learning (DL) methodology is utilized to train a model employing a support vector machine and a Convolution Neural Network (CNN) for extracting the dominating features and detect breast cancer from test image samples. Using the dataset of the tissue cells utilized in the test samples, the automatic detection model divides the mammography images into ‘Malignant’ and ‘Benign’ breast cancers. This study compares the accuracy of the findings for identifying breast cancer for several kernels, such as sigmoid function, radial basis function, polynomial function, and linear function. With an accuracy of 95.78%, the linear kernel outperforms others. The training set of data contains a variety of tissue cell samples, and their recognition performance are tested until it exceeds future expectations.