Authors:
Keisuke Negoro
;
Takeshi Nakazaki
;
Susumu Takeuchi
and
Masanori Akiyoshi
Affiliation:
Graduate School of Information Science and Technology, Osaka University, Japan
Keyword(s):
Qualitative and Quantitative Simulation, Monte Carlo Method, Inverse Propagation, Contradiction.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Enterprise Software Technologies
;
Intelligent Problem Solving
;
Knowledge Engineering and Ontology Development
;
Knowledge-Based Systems
;
Software Engineering
;
Symbolic Systems
Abstract:
In order to decide an effective management plan, managers often draw up and evaluate business scenarios. To make the evaluation, a simulation method on the qualitative and quantitative hybrid model represented as causal graph has been proposed. There is a strong need to get optimal input values for the target outputs in the simulation, but exhaustive search can not be realistically applied to it from considering the processing time. Therefore, we propose a quick search method for optimal input values cencerning the qualitative and quantitative hybrid simulation. Our approach is to get optimal values of input nodes by inverse propagation of effects from the value of target output nodes on the simulation model. However, it generates the contradiction that the value of a separated node in the causal graph decided from one of destination nodes is different from the value of the other destination nodes. Therefore, we re-execute the inverse propagation repeatedly from the nearest qualitati
ve node connecting to a quantitative node for solving the contradiction. By experimental results about the proposed method, time could be reduced for reaching the solution. We also could confirm a certain level of accuracy about the solution.
(More)