Building a Movie Recommendation system

The following report entails details on a data science project that I took up as part of the HarvardX PH125.9x Data Science: Capstone final course, about creating a machine learning algorithm for a movie recommendation system. The algorithm is expected to have an RMSE value of 0.86490 or lesser, implying good accuracy. The creation of this algorithm was originally tasked to data scientists across the globe, as part of the Netflix Challenge 2006. The dataset used for this analysis was the 10M version of the Movielens dataset, the links of which were accessed directly from the edx course. The given dataset was split into 2 sub-datasets called the edx and validation sets. The edx set was further sub-divided into the test and training sets and the algorithm was built on these datasets accordingly. Finally, the final model was tested on the validation dataset and its RMSE was recorded.