Optimizing ship air-defense evaluation model using simulation and inductive learning

Abstract

This thesis presents an effective method to integrate simulation modeling with inductive learning the analyze ship air-defense combat scenarios. By combining the use of inductive learning with simulation, we are able to discover rules in a ship air-defense evaluation model about the optimal weapon assignments that we might not be aware of or could not express clearly. This approach can also perform sensitivity analysis in identifying variables that are critical for certain weapons operations. In addition, results from inductive learning, as represented in the format of decision trees, are easy for a human user to understand, maintain and adopt for other use.