Application of machine learning in recommendation systems

Study includes basic information about machine learning and recommender systems with their examples. More broadly addressed was the topic of machine learning's algorithms, which are used in such systems. The paper mainly focused on filtering algorithms based on the neighborhood of users or objects, and based on content. The description of these algorithms includes: similarities, disadvantages and advantages, measures for evaluating the algorithm, and calculation of the sample value of the evaluation prediction. The design part of the work begins with the description of the used databases from the MovieLens portal. Afterwards, the technology and practical implementation of the algorithms described above are then presented. The next part contains an analysis of the results and conclusions based on the simulations carried out on the computer to assess how the algorithms work. At the end of the work, there is a summary, performance evaluation of recommendation systems, and lessons learned from the project, as well as a proposal for further work on the issue of such systems.