Breast cancer remains a critical health problem facing throughout the world, requires accurate and timely detection for effective treatment and improved patient outcomes. Recently, ML and DL methods have drawn a lot of interest for their potential to enhance breast cancer detection and diagnosis. This paper offers a thorough evaluation of the most recent findings in the field of ML and DL-base breast cancer diagnosis. The study further presents the union of existing studies that highlights the significant features in this field, while also identifying areas for future research, includes the combination of recent technologies, real-world validation, and clinical adoption.