In the previous chapter, we focused on supervised learning for predicting continuous numerical values using regression. Now, we shift our attention to another major category within supervised learning: classification. The objective in classification is different. Instead of predicting a number, we aim to assign an input data point to one of several predefined categories or classes.

This chapter introduces core concepts and algorithms for tackling classification problems. You will learn:

• What defines a classification problem and how it differs from regression.
• The fundamentals of Logistic Regression, a common algorithm for binary classification tasks, including the role of the sigmoid function and the concept of decision boundaries.
• An alternative approach using the K-Nearest Neighbors (KNN) algorithm, an instance-based learning method.
• Basic techniques for evaluating the performance of classification models, moving beyond simple accuracy.
• Practical steps to implement a KNN classifier.