Abstract
This chapter is devoted to the study of algorithms for construction of tests, decision rules and trees. Our aim is to construct tests with minimum cardinality, decision rules with minimum length, and decision trees with minimum depth. Unfortunately, all the three optimization problems are NP-hard. So we consider not only exact but also approximate algorithms for optimization.
The chapter consists of four sections. In Sect. 4.1, we study approximate (greedy) algorithms for optimization of tests and decision rules. These algorithms are based on greedy algorithm for the set cover problem.
Section 4.2 deals with greedy algorithm for decision tree construction.
In Sect. 4.3, we study exact algorithms for optimization of decision trees and rules which are based on dynamic programming approach. We show that if P ≠ NP then there is no similar algorithms for test optimization.
Section 4.4 contains conclusions.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Rights and permissions
Copyright information
© 2011 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Moshkov, M., Zielosko, B. (2011). Algorithms for Construction of Tests, Decision Rules and Trees. In: Combinatorial Machine Learning. Studies in Computational Intelligence, vol 360. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-20995-6_4
Download citation
DOI: https://doi.org/10.1007/978-3-642-20995-6_4
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-20994-9
Online ISBN: 978-3-642-20995-6
eBook Packages: EngineeringEngineering (R0)