Apply bagging, boosting, and random forests to a data set of your choice (not one used in the committee machines labs). Fit the models on a training set, and evaluate them on a test set. How accurate are these results compared to more simplistic (non-ensemble) methods (e.g., logistic regression, kNN, etc)? What are some advantages (and disadvantages) do committee machines have related to the data set that you selected?

Any data set you want to use or have access to.