This chapter explores the fundamental differences between Supervised and Unsupervised Learning, two important families of algorithms in the field of Machine Learning. The basic methodologies and distinctions between these paradigms are presented and described regarding model optimization and evaluation. In Supervised Learning, classification and regression problems are introduced. Classification problems attempt to learn from known categories in the data, while regression problems involve a continuous numerical variable as the outcome of the prediction. On the other side, Unsupervised Learning is presented as a different paradigm where the discovery and learning from data is performed without prior information or known labels but by exploring the hidden structures of the dataset. Furthermore, this chapter also documents the different evaluation challenges and techniques. The utilization of these different paradigms depends on the business problems and the data requirements at hand. However, having a broad overview of the possible methodologies available and how they interrelate is crucial for sustainable and reliable Machine Learning integration.