zbMATH — the first resource for mathematics

Data mining with decision trees. Theory and applications. (English) Zbl 1154.68098
Series in Machine Perception and Artificial Intelligence 69. Hackensack, NJ: World Scientific (ISBN 978-981-277-171-1/hbk; 978-981-277-172-8/ebook). xviii, 244 p. (2008).
This book covers an important area in the application of computer-based methods to the discovery, formalization and application of existing knowledge that can and has to be transferred to computers. Very often this material exists in the format of huge data sets, possibly already on the computer in the format of databases, but also still on paper. Think, for instance, about the huge amount of medical or legal cases collected over a very long time.
Decision trees are one important means in solving this problem. They are a well explored possibility and technology for an efficient and accurate decision process based on a sequence of single tests. Originally they are a data structure for the representation and the use of logic functions, at present the formal logical background is not necessarily used explicitly, but remains as a well understood feature in the background.
This book is a comprehensive collection of “everything” that relates to this process, and it is nice to have all the relevant topics at one place. It consists of three parts:
Part 1: Foundations, Part 2: Algorithms, Part 3: Advanced topics.
The different chapters illustrate very well the content of this book: Introduction to decision trees; Growing decision trees; Evaluation of classification trees; Splitting criteria (for the splitting of nodes); Pruning trees; Advanced decision trees; Decision forests; Incremental learning of decision trees; Feature selection; Fuzzy selection trees; Hybridization of decision trees with other techniques; Sequence classification using decision trees.
Many algorithms are given in pseudocode and can be used for further research and implementations.
In general, the book is a very useful and nice coverage of the field. It would have been desirable, however, also to cover the area of small example sets, which is a very general problem in many practical applications. It is highly recommendable for people who want to begin working in this field and need guidance to start into the large area of applying these methods.

68T05 Learning and adaptive systems in artificial intelligence
68-01 Introductory exposition (textbooks, tutorial papers, etc.) pertaining to computer science
68P05 Data structures
68P15 Database theory
Full Text: DOI