Statistics and data mining often get bundled together, but (in my opinion), they're generally different practices with different goals. As a language designed for statistics, much of R's core functionality is focused on exploring and understanding data: model design, inference, and visualization. But when your goal is simply to get the best predictions from a big data set (without worrying too much about the model itself), much of R's statistical power can also be put to data mining purposes.
