You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
yadavamrendra edited this page Nov 13, 2017
·
2 revisions
Abstract:
It is shown that the system compares favorably with commonly used symbolic learning methods that use heuristic rather than explicit map methods to guide their search in the rules space. First-order learning involves finding a clause-form definition of a relationship based on examples of the relationship and relevant contextual information. In this article, a particular first-order learning system is modified to customize it to find functional relationship definitions. This restriction leads to faster learning times and, in some cases, definitions that have greater predictive accuracy. Other first-class learning systems could benefit from similar specialization. This article describes FORE, a system that learns horn clauses from data expressed as relationships [2]. FOIL is based on ideas that have proven effective in attribute-value learning systems, but extends them to first-rate formalism. This new system has been successfully applied to several tasks from the machine learning literature.