This report analyzes data collected from accelerometer sensors on the belt, arm, forearm, and dumbbell while subjects performed barbell lifting exercises. A machine learning model is trained to predict the manner in which the exercise was performed (classe). Data preprocessing, cross-validation, model selection, and expected out-of-sample error are discussed. The final random forest model is applied to 20 test cases for submission.

Milestone report for the SwiftKey Capstone Project. Includes exploratory data analysis, dataset summaries, and plans for building the next-word prediction model and Shiny application.

developing data products