Furkan Çınar, Root Cause Analysis with Probabilistic Graphical Bayes Network Models Integrated with Affinity Diagrams

M.S. Candidate: Furkan Çınar
Program: Information Systems
Date: 29.07.2025 / 10:30
Place: 
A-212

Abstract: Qualitative management tools are widely used to analyze root causes of management problems in the industry. However, the use of these tools for business problems are challenging. Their application requires considerable time and effort, their analysis are open to personal interpretation and their results may lack quantitative and structured data. This study proposes a novel approach that combines a qualitative management tool and a quantitative modelling tool to analyze the root causes of problems encountered in the business world. This approach makes it possible to use both qualitative and quantitative data and to include the human factor in the process while performing root cause analysis. The proposed approach is not specific to a specific type of industry, they can be applied to any management problem where a group of domain experts are available. The proposed approach combines the Affinity Diagram, which is one of the 7 Management Tools, with a probabilistic modelling approach called Bayesian Networks. Qualitative data and domain knowledge provided by domain experts who have in-depth knowledge about the problem are systematically used with Affinity Diagrams to create a Bayesian Network model. This study demonstrates the application of this approach and its results to a real-world problem in production of electronic components.