Sistem penunjang keputusan kelayakan pemberian pinjaman dengna metode fuzzy tsukamoto

Decision support systems (DSS) can be used to help settlement issues or decisions that are semi-structured or structured. The method used is Fuzzy Tsukamoto. PT Triprima Finance is a company engaged in the service sector lending with collateral in the form of Motor Vehicle Owner Book or car (reg). PT. Triprima Finance should consider borrowing from its customers with the consent of the head manager. Such approval requires a long time because they have to pass through many stages of the reporting procedure. Decision-making activities at PT Triprima Finance carried out by the analysis process manually. To help overcome these problems, the need for completion method in accuracy and speed of decision making feasibility of lending. To overcome this need to develop a new system that is a decision support system Tsukamoto fuzzy method. is expected to facilitate kaposko to determine the decisions to be taken.

💡 Research Summary

This paper presents the design, implementation, and evaluation of a decision‑support system (DSS) for loan‑eligibility assessment at PT Triprima Finance, a mid‑size Indonesian finance company that provides vehicle‑title‑secured loans. The existing manual workflow requires multiple hierarchical approvals and typically consumes three to five days per application, leading to delays, inconsistent judgments, and high operational costs. To address these shortcomings, the authors adopt the Fuzzy Tsukamoto inference method, which is particularly suited for problems where the output must be a continuous “degree of suitability” rather than a crisp yes/no decision.
The study begins with a literature review that highlights the limited use of the Tsukamoto model in financial credit scoring, despite its advantages over Mamdani and Sugeno approaches for generating direct possibility values. The authors then identify four key input variables—monthly income, debt‑to‑income ratio, credit score, and collateral value (the market value of the motor vehicle). Each variable is fuzzified using three linguistic labels (Low, Medium, High) represented by triangular membership functions normalized to the