In this chapter, we discussed what recommender systems are and the types of these that exist today. We studied two main approaches to building recommender systems: content-based recommendations and collaborative filtering. We identified two types of collaborative filtering: user-based and item-based. We looked at the implementation of these approaches, and their pros and cons. We found out that an important issue in the implementation of recommender systems is the amount of data and the associated large computational complexity of algorithms. We considered approaches to overcome computational complexity problems, such as partial data updates and approximate iterative algorithms, such as ALS. We found out how matrix factorization can help to solve the problem with incomplete data, improve the generalizability of the model, and speed up the calculations. Also, we implemented a system of collaborative filtering based on the linear algebra library and using the mlpack general-purpose machine learning library.

It makes sense to look at new methods such as autoencoders, variational autoencoders, or deep collaborative approaches applied to recommender system tasks. In recent research papers, these approaches show more impressive results than classical methods such as ALS. All these new methods are non-linear models, so they can potentially beat the limited modeling capacity of linear factor models.

In the next chapter, we discuss ensemble learning techniques. The main idea of these types of techniques is to combine either different types of machine learning algorithms or use a set of the same kind of algorithms to obtain better predictive performance. Combining a number of algorithms in the one ensemble allows us to get the best characteristics of each one, to cover disadvantages in a single algorithm.