Files for this blog post: https://moritzkoerber.github.io/analysis/r/2019/07/24/blogpost/
Going to the gym, whether to boost your health, to lose weight, or simply because it is fun, is certainly a worthwile activity. FiveThirtyEight recently reported that according to the latest Physical Activity Guidelines for Americans, every form of activity counts. However, if you have eager goals, not only quantity but also quality and being efficient matters. In this project, I predict whether a certain exercise, a barbell lift, was well or sloppily executed on the basis of data obtained from accelerometers on the belt, forearm, arm, and dumbell of six participants. The participants performed the exercises correctly and incorrectly in five different ways.
The idea for this analysis stems from the Coursera course Practical Machine Learning by Johns Hopkins University. The data for this project come from http://groupware.les.inf.puc-rio.br/har.