Neural Random Forests

Decision tree learning is a popular data-modeling technique that has been around for over fifty years in the fields of statistics, artificial intelligence, and machine learning. The approach and its innumerable variants have been 1 successfully involved in many challenges requiring classification and regression tasks, and it is no exaggeration to say that many modern predictive algorithms rely directly or indirectly on tree principles. What has greatly contributed to this success is the simplicity and transparency of trees, together with their ability to explain complex data sets. The monographs by Breiman et al. (1984), Devroye et al. (1996), Rokach and Maimon (2008), and Hastie et al. (2009) will provide the reader with introductions to the general subject area, both from a practical and theoretical perspective. The history of trees goes on today with random forests (Breiman, 2001), which are on the list of the most successful machine learning algorithms currently available to handle large-scale and high-dimensional data sets.