Efficient Relational Learning from Sparse Data

  • Authors:
  • Lubos Popelínsky

  • Affiliations:
  • -

  • Venue:
  • AIMSA '02 Proceedings of the 10th International Conference on Artificial Intelligence: Methodology, Systems, and Applications
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

This work deals with inductive inference of logic programs -relational learning - from examples. The work is, in the first place, application-oriented. It aims at building an easy-to-use relational learner and it focuses on the tasks that are solvable with the tool. Assumption-based learning, the new learning paradigm is introduced and the ABL system WiM is described. A methodology for experimental evaluation of ILP systems is introduced and experiments with WiM are displayed. Two classes of application - database schema redesign and mining in spatial data - that have been successfully solved with WiM are described.