University of Twente Student Theses
An Integrated Process Mining and Data Mining Approach for the Validation of Agent-based Simulation Models
Govers, Ruben (2023) An Integrated Process Mining and Data Mining Approach for the Validation of Agent-based Simulation Models.
This is the latest version of this item.
|
PDF
1MB |
| Abstract: | Agent-based simulations and process mining are upcoming fields. However, literature research found that research on agent-based models regularly lacks proper validation or an understanding of this validation. Moreover, the bounds of input parameters in which a model operates validly are often unknown either. Uncertainty regarding this might cause an agent-based model to become invalid as it gets adjusted over time. This thesis presents an innovative approach for validating agent-based simulation models based on a three-cycle interpretation of CRISP-DM, a renowned data mining methodology. Each cycle aligns with a separate discipline. The first involves agent-based simulation, the second process mining, and the third data mining. The thesis follows the design science research methodology to deliver the approach. As part of this methodology, four objectives were identified. The approach should be interdisciplinary, adaptable, effective, and interpretable. The approach was demonstrated and evaluated through a detailed case study of the Schelling model of segregation. The input parameters that influence the model in the case study include topology, heterogeneity, and agent rules. The case study shows that the approach achieved the objectives within the research context. The findings highlight the potential of this integrated approach across a range of fields, creating a new path for further advancements in the field of agent-based models. Examples of such advancements include automatic agent-based model generation from event logs or the creation of interpretable agent-based models using data mining. |
| Item Type: | Essay (Master) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Business Information Technology MSc (60025) |
| Link to this item: | https://purl.utwente.nl/essays/96531 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page