Title :
| Presenter | Ali El Hassouni |
| Abstract |
Title : Learning Datalog Programs from Input and Output
| Presenter | Xu Wang |
| Abstract | A new inductive learning task is proposed for Datalog programs. In the learning task an example is a pair hI;Oi where I and O, standing for input and output respectively, are sets of ground atoms. It means that O is the least (Herbrand) model of a learned Datalog program when it is given the input I. An inductive learning algorithm is presented for this inductive learning task. It is a modular and anytime algorithm. |