Breast cancer is a prevalent and significant form of cancer that has a profound impact on women worldwide, and it is the leading cause of mortality. Timely identification of breast cancer is a crucial procedure that can enable suitable therapy, hinder the advancement of cancerous cells, and decrease rates of illness and death. Machine Learning (ML) is highly favored techniques for precise detection and classification of breast cancer. This paper introduces a unique method for categorizing breast cancer by utilizing a novel technique for selecting relevant attributes and employing ML algorithms. This method is referred to as the Breast Cancer Detection Model (BCDM). The BCDM system comprises three distinct phases, namely data preprocessing, feature selection, and patient detection. The primary objective of the data preprocessing phase is to extract distinctive characteristics from mammogram images and eliminate any anomalous items. This is accomplished by employing an innovative feature selection technique known as Enhanced Bat Algorithm (EBA) to identify the most influential and meaningful attributes from the extracted ones. Afterwards, these attributes are entered into the patient detection phase to determine whether the patient is unaffected or has either benign or malignant breast cancer. Based on the obtained results, the proposed method surpasses rivals in performance in terms of accuracy, precision, recall, and F-score.