Cancer remains one of the leading causes of mortality worldwide, necessitating early and accurate detection mechanisms to improve patient survival rates. Traditional diagnostic methods, while effective, often face challenges regarding time efficiency, inter- observer variability, and sensitivity. In recent years, Machine Learning (ML) and Deep Learning (DL) have emerged as pivotal tools in oncology, offering automated, high-precision diagnostic capabilities. This paper reviews the strengths of various ML paradigms—including Support Vector Machines (SVM), Random Forests (RF), and Convolutional Neural Networks (CNN)—in the detection of malignancies. We critically analyze the performance of these algorithms across different cancer modalities, such as breast, lung, and skin cancer. Furthermore, the review highlights the transition from feature-based classical ML to automated feature extraction via Deep Learning, discusses current challenges such as data heterogeneity and model interpretability, and proposes future directions for integrating AI into clinical workflows.