Decision forests: a unified framework for classification, regression, density estimation, manifold learning and semi-supervised learning. (English) Zbl 1243.68235

Criminisi, Antonio; Shotton, Jamie; Konukoglu, Ender

Found. Trends Comput. Graph. Vis. 7, No. 2-3, 81-227 (2011).

The book presents a decision forest framework that encompasses classification, regression, density estimation, manifold learning and semi-supervised learning under the same roof. A general core was first developed and the different types of learning can be further instantiated from that in application to various tasks, ranging from scene and object recognition, automated medical diagnosis and semantic text parsing.

Many types of readers can enjoy this very interesting and practical book: from the students willing to know the foundation of decision forests and researchers updating their knowledge with new contributions to the field to practitioners working in the applicative fields targeted by this book. It can even be fascinating just for gratifying one’s curiosity as to how is it that Kinect works so nice for Xbox 360.

Reviewer: Catalin Stoean (Craiova)

MSC:

68T05	Learning and adaptive systems in artificial intelligence
68-02	Research exposition (monographs, survey articles) pertaining to computer science
62H30	Classification and discrimination; cluster analysis (statistical aspects)
62G07	Density estimation

Keywords:

decision forests; machine learning; computer vision; classification; regression; density estimation; manifold learning; semi-supervised learning

Software:

SHOGUN; SVMlight