Extensions to the Relational Paths Based Learning Approach RPBL
Zhiqiang Gao, Zhizheng Zhang, Zhisheng Huang · 2009
In this paper we extend RPBL, a Relational Paths Based Learning approach for first order theories in three directions. We apply domain theories to expand structured instance space, learn recursive theories by an example of learningmember relationship of lists, and analyze the performance as well as time complexity theoretically. In addition, we give the details of our experimental results.