Agent-based simulation and artificial intelligence grounded management information system: The application of Anadolu University open education system

In social systems, the decisions made by the administrators have the potential to affect the system deeply. Therefore, there is a need for various instruments that provide more information to decision makers and facilitate decision making processes. Thanks to the developing technology, it has become possible to conduct experiments on these social systems by using agent-based simulation methods. A simulation model that better represents real life can be obtained by integrating artificial intelligence, which is being recently used in many fields, with agent-based simulation. In this study, a modelling approach has been developed that includes the use of the multilayer perceptron, which is one of the artificial intelligence techniques, and the agent-based simulation. Basically, it is aimed to compose a management information system that uses this approach and allows to test various scenarios and decisions. This management information system is implemented specifically for Anadolu University Open Education System and modelling is done by collecting data from 41,000 students registered in the system. Since the proposed modelling approach is aimed to be usable for many social systems with similar structures, the study contains unique points in many respects and includes data, analysis processes and methods that can set an example for studies that can be done in similar fields. The validation and verification stages of the developed simulation model and the effects of the changes that may occur in the internal and external variables on the system in the long and short terms are presented. In addition, four basic scenarios that demonstrate the capability of the developed management information system and can serve as an example for system users are simulated.