Fundamental Problems in Machine Learning

Classification: This problem type entails a program determining the class or label of a data point among C different labels. For instance, consider the task of classifying a picture of handwritten numbers with ten labels representing the digits 0 to 9. In this scenario:

– Task: Identifying the label of an image depicting handwritten digits.

– Evaluation: The number of correctly classified images.

– Experience: The dataset comprises pairs of (number images, labels) known in advance.

Regression: These problems involve the model producing continuous values as output. The task of predicting real estate prices is a regression problem.

Clustering: These problems require the division of data points into small clusters based on the relationship between the data within each cluster. For instance, consider customer clustering based on purchasing behavior.