This report describes different timeseries and machine learning forecasting models applied to a real stock close price dataset. The personal HarvardX: PH125.9x: Data Science: Capstone Movielens Project encourage us to apply machine learning techniques that go beyond standard linear regression. For this project we will start with a general idea of the stock price, incluiding dataset analysis. Followed by a general description and analysis of the dataset, our objective is to apply different forecasting predictive models for “Amazon” stock daily close price. The models will be evaluated, analysed and compared, following the main course project directions. For this evaluations we will focus in the train set accuracy performance of the models, trying to simplify our analysis due to our aim is showing new forecasting applications and encourage the open source use of them. To apply our models, the dataset will be downloaded using the function getsymbols() from quantmod package. The data will be prepared to predict the next 30 days close price from today. The results will be explained during the report and concluding remarks.