This repository contains my solution to the first practical assignment of Recommender Systems class of Departamento de Ciência da Computação - Universidade Federal de Minas Gerais.

The solution uses Item-Based Collaborative Filtering Techniques with mean centering normalization and cosine similarity.

If you have to develop a similar assignment do not use this job as if you were the author, plagiarism is a very bad practice. You can use the code/documentation to get insights into your problem.

About the files:

• main.cpp - Full code (I know that is a bad practice to write single-file codes, but I have some problems managing my time and I had to develop all code in a small amount of time).
• Makefile - Compiling commands
• ratings.csv - Train data
• targets.csv - Test data
• submission.csv - My output for test data using the train data
• doc.pdf - Documentation of assignment (In Portuguese)