This repository hosts the source code of the Machine Learning Coursera assignment
Using devices such as Jawbone Up, Nike FuelBand, and Fitbit it is now possible to collect a large amount of data about personal activity relatively inexpensively.
This document will describe the analysis performed on a dataset containing 5 classes of physical exercises collected over a period of 8 hours for multiple subjects. Details of the experiment can be found at this link.
We will first load the data locally and clean it up of invalid observation. The data is composed of:
- Training set dataTrain: 70% of which will be used to create the model and 30% will be used for testing the model before applying it to the given testing set
- Testing set dataTest
We will apply a “RandomForest” method to build our prediction. At the end we will run the model on the testing set.