03:30 PM - 05:10 PM
Random forests (Breiman, 2001), that are now part of the essential toolbox of any data analyst, is a very powerful non-parametric statistical learning method that can be used for classification, regression and many other problems including survival data. One of its main appeal is its ability to effectively capture nonlinear dependencies and interactions. We will discuss the basic properties of a random forest, review the useful implementations, some of the many extensions to more complex settings, and the available theoretical results. Random forests are still a very active area of research and we will provide an overview of current topics and recent developments.