"Monotone Boolean Function Learning Techniques Integrated with User Interaction"
Proceedings of the Workshop on Learning from Examples
which took place during the 12-th Inter'l Conference on Machine Learning,
Lake Tahoe, Calif., July 9-12, pp. 41-48, 1995.
by Kovalerchuk, B., E. Triantaphyllou, and E. Vityaev
Abstract:
This paper discusses some key issues about an interactive learning approach
based on the theory of monotone Boolean functions. We present some key problems
and the main steps of some algorithms for solving them. The concept of Shannon functions
is used as a criterion for algorithmic optimality. The proposed approach allows for the
possibility to decrease the number of positive and negative examples needed to infer a monotone
Boolean function, and thus can make this type of machine learning systems more applicable.
Key Words:
Learning from examples, Monotone Boolean Functions, Shannon Function.
(size = 1,030 KB)
