Iterative learning for multiple phases planning: phased-REPLE

In a status selection planning system, which is a kind of knowledge based planning system, quality of the solution depends on the status selection rules. However, it is usually difficult to acquire useful knowledge from human experts. We propose an iterative learning method of a status selection rule using inductive learning. The planning process is divided into stages. Then, a phase is a bundle of stages. Status selection rules for phases are acquired from the training set which has been gathered from each phase from the last, phase. The rules are used to gather training sets of the next iteration. The proposed method is applied to a job shop problem.