My data analysis of the Samsung dataset for the Data Analysis course taught by Jeff Leek on Coursera.
Here I attempt to use combine SVD for dimensionality reduction and SVM for classification in order to predict a user's activity (e.g. standing, sitting, walking) based on smartphone accelerometer and gyroscope data.
View my report here.
To reproduce the exact analyses presented in the report, run source('00-download.r') to get the dataset then source('main.r') from the R command line.
- reshape
- ggplot2
This work is licensed under the Creative Commons Attribution-ShareAlike 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-sa/3.0/ or send a letter to Creative Commons, 444 Castro Street, Suite 900, Mountain View, California, 94041, USA.