Ever wondered how can we predict the gasoline price in upcoming months? How the projected exchange rates of currencies are determined? The crux of these problems is the ability to predict a value in a continuous range. The algorithms that solve those problems are called regression algorithms. The name regression suggests that these algorithms are mostly iterative in nature. This is different than classification because in classification we need to predict either one of the two values (in case of binary classification) or one of the many (a set of finite labels, in case of multiclass classification) labels. On the other hand, in these situations the predicted value will have to be real value and that’s regression. In this chapter, you shall learn about several types of regression algorithms that ML.NET provides and how to measure performance of these algorithms. In some literature, these algorithms are termed as curve fitting algorithms.