Learning by Discovering Conflicts

Abstract

The paper describes a novel approach to inductive learning based on a ‘conflict estimation based learning’ (CEL) algorithm. CEL is a new learning strategy, and unlike conventional methods CEL does not construct explicit abstractions of the target concept. Instead, CEL classifies unknown examples by adding them to each class of the training examples and measuring how much noise is generated. The class that results in the least noise, i.e., the class that least conflicts with the given example is chosen as the output. In this paper, we describe the underlying principles behind the CEL algorithm, a methodology for its construction, and then summarize convincing empirical evidence that suggests that CEL can be a perfect solution in real-world decision making applications.