My Shiny App is a deployment of my final project from class Practical Machine Learning. Remember, in that project we developed a model to predict wheher a person was making a movement in a proper way (outcome A) or was making one of five common mistakes (outcome B,C,D,E).

• Cleaning data: Remove predictors which contain almost NA values.
• Select predictors: Applying a correlation criteria I removed highly correlated predictors.
• Train and test diferent models (Regresions, Boosting, knn, random forest…)
• Final model: Random Forest with 32 predictors reach 0.99 accuracy in the test data and Quiz data.
• The whole project and code can be seen here

knitr::include_graphics("introtab.jpg")

In tab this you’ll find a summary of the problem and quick instructions about the use of the App.

knitr::include_graphics("apptab.jpg")

Just Pick a number beetwen 1 and 20 and see the predicted classe!