|Title||Bottom-up induction of feature terms|
|Publication Type||Journal Article|
|Year of Publication||2000|
|Authors||Armengol E, Plaza E|
|Journal||Machine Learning Journal|
The aim of relational learning is to develop methods for the induction of descriptions in representation formalisms that are more expressive than attribute-value representation. Most work on relational learning has been focused on induction in subsets of first order logic like Horn clauses. We think that Machine Learning research can also profit from exploring other representation formalisms that facilitate the expressivity of relations but are different subsets of first order logic. In this paper we introduce a representation formalism based on feature terms and INDIE, a bottom-up learning method that induces class descriptions in the form of feature terms from positive and negative examples. INDIE is based on the subsumption and anti-unification operations over feature terms.
- About IIIA
- Current news