Loading...

Application of Artificial Intelligence in Improving Multi-Criteria Decision-Making Methods

Mokhtari Torkaman, Hossein | 2025

73 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 58381 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Eshghi, Kourosh
  7. Abstract:
  8. In the current era—defined by data and artificial intelligence—the integration of AI-based tools into diverse applications and decision-making contexts has grown markedly. In particular, opportunities exist to automate the construction of pairwise comparison matrices using empirical data and machine learning algorithms. In this paper, a data-driven decision-making framework is proposed, comprising three sequential phases. In the first phase, the decision maker (DM) employs two machine learning techniques—Principal Component Analysis (PCA) and Random Forest—to identify and select the most relevant decision criteria. In the second phase, once non-essential criteria have been removed from the dataset, another machine learning model—the Bradley-Terry model, which belongs to the preference learning subdomain—is utilized to uncover decision-maker preferences and generate pairwise comparisons among the selected criteria. In the third phase, the extracted comparisons are then fed into an MCDM method to complete the decision-making process. The proposed approach was applied to a case study involving mobile phone selection, using a real-world dataset containing specifications and features of various smartphone models. The results demonstrated high-quality decision outcomes, highlighting the effectiveness of the framework. Overall, the study confirms that AI and machine learning techniques can significantly enhance both the efficiency and accuracy of preference learning and criteria selection in pairwise comparison-based MCDM methods
  9. Keywords:
  10. Artificial Intelligence ; Machine Learning ; Decision Making Methods ; Multicriteria Decision Making ; Pairwise Comparisons

 Digital Object List

 Bookmark

No TOC