XCS for Missing Attributes in Data

To handle missing attributes when acquiring knowledge from data in real-world problems, this paper proposes XCS for Missing Attributes in Data (XCS-MA). XCS-MA complements the value of the missing attributes of the input data and estimates the correct output for its complemented input. If the correct output is the same for all possible complemented inputs, XCS-MA uses the missing attribute data for the learning. In experiments, XCS-MA correctly estimated the accuracy of the classifiers (represented by the if-then rules) and acquired the optimal classifier subset in the 11-Multiplexer problem with the missing attributes. Once the optimal classifier subset is acquired, XCS-MA selects the correct output for any input even if many missing attributes are included in the data.

• Publication year

  2018

• Venue

  2018 Joint 10th International Conference on Soft Computing and Intelligent Systems (SCIS) and 19th International Symposium on Advanced Intelligent Systems (ISIS)

• Publication date

  2018-12-01

• Fields of study

  Computer Science

• Identifiers
  DOI 10.1109/SCIS-ISIS.2018.00063
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• XCS-CR: determining accuracy of classifier by its collective reward in action set toward environment with action noise

  2018cited by this paper
• Toward a theory of generalization and learning in XCS

  2004cited by this paper
• The CN2 induction algorithm

  2004cited by this paper
• A Comparison of Several Approaches to Missing Attribute Values in Data Mining

  2000cited by this paper
• An algorithmic description of XCS

  2000cited by this paper
• An Analysis of Generalization in the XCS Classifier System

  1999cited by this paper
• Credit Assignment for Pole Balancing with Learning Classifier Systems. In

  1995influential reference
• Escaping brittleness: the possibilities of general-purpose learning algorithms applied to parallel rule-based systems

  1995cited by this paper
• Genetic Algorithms in Search Optimization and Machine Learning

  1988cited by this paper
• Learning to predict by the methods of temporal differences

  1988cited by this paper

• No citing papers are available for this paper.