Breast cancer (BC) is the leading cause of death for women worldwide, both in industrialised and developing nations. Breast cancer (BC) is a type of cancer that develops in the breast cells and is characterised by gene alterations, persistent pain, changes in size, a reddish colour, and altered skin texture. A variety of imaging modalities, including magnetic resonance, mammography, and ultrasound, are often used to diagnose BC. Mammography and ultrasound image experimentation will lead to low sensitivity and specificity in recognising lesions and distinguishing between benign and malignant lesions. A malignant lesion is cancerous, while a benign lesion is not. The patient will therefore need to undergo expensive bioptics operations. In this study, we examined other metrics like accuracy, precision (Pr), and F-score to help with breast cancer detection. We aggregated these deep features into a single set of three hundred features. Each of these deep models has the ability to extract 1 K deep features, which are subsequently reduced to hundred deep features using the recursive feature elimination (RFE) approach. And used these 300 features as input for decision trees (DT), support vector machines (SVM), linear regression, k-nearest neighbourhood (kNN), and linear discernment analysis classifiers. With an accuracy of 97.22, the SVM machine classifier produced the most effective results out of all of these classifiers.