A Support Vector Machine (SVM) is a supervised learning algorithm used for classification and regression tasks. It identifies decision boundaries by maximizing the margin between data points, making it effective in text classification, image recognition, and anomaly detection.