Perancangan Kualitas Sistem Pendeteksi Transaksi Mencurigakan Pada Mobile Banking Menggunakan Pendekatan Quality by Design

Neysa, Amanda Na'ilah (2026) Perancangan Kualitas Sistem Pendeteksi Transaksi Mencurigakan Pada Mobile Banking Menggunakan Pendekatan Quality by Design. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5010221186-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Perkembangan layanan mobile banking meningkatkan kemudahan transaksi keuangan sekaligus memperbesar risiko terjadinya transaksi mencurigakan (fraud) yang semakin kompleks, sehingga diperlukan sistem deteksi yang lebih adaptif dibandingkan metode konvensional. Penelitian ini mengembangkan sistem deteksi transaksi mencurigakan menggunakan pendekatan Quality by Design (QbD) yang diintegrasikan dengan machine learning melalui tahapan identifikasi kebutuhan sistem, pengolahan data (data cleaning, feature engineering, dan penanganan imbalanced dataset), serta pengembangan model klasifikasi menggunakan Support Vector Machine (SVM) dan Random Forest. Pendekatan QbD diterapkan melalui penetapan Quality Target Product Profile (QTPP), identifikasi Critical Quality Attributes (CQA), penentuan Critical Process Parameters (CPP), serta analisis Design of Experiments (DoE) dan General Linear Model (GLM) untuk menentukan design space sistem. Hasil penelitian menunjukkan bahwa SVM memberikan performa terbaik dalam mendeteksi fraud dengan nilai Recall 5,76% dan F1-Score 0,0403, lebih tinggi dibandingkan Random Forest yang memperoleh Recall 0,62% dan F1-Score 0,0104. Analisis QbD mengidentifikasi lima parameter kritis, yaitu Amount, Location, Interval, Device, dan Receiver Account, serta menunjukkan bahwa interaksi antarparameter memberikan pengaruh yang lebih signifikan terhadap fraud rate dibandingkan pengaruh masing-masing parameter secara individual. Penelitian ini menghasilkan design space yang dapat menjadi dasar dalam proses monitoring transaksi mencurigakan dan mendukung pengembangan sistem deteksi fraud yang lebih sistematis, adaptif, dan sesuai dengan kebutuhan operasional perbankan.
=======================================================================================================================================
The rapid growth of mobile banking services has improved the convenience of financial transactions while simultaneously increasing the risk of increasingly complex and dynamic fraudulent transactions. Conventional fraud detection systems have limitations in identifying evolving fraud patterns, highlighting the need for a more systematic and data-driven approach. This study develops a suspicious transaction detection system for mobile banking by integrating the Quality by Design (QbD) approach with machine learning techniques. The research methodology includes system requirement identification through literature review and interviews with banking fraud detection analysts, data preprocessing consisting of data cleaning, feature engineering, and imbalanced dataset handling, followed by the development of classification models using Support Vector Machine (SVM) and Random Forest. The QbD framework was implemented by defining the Quality Target Product Profile (QTPP), identifying Critical Quality Attributes (CQA), determining Critical Process Parameters (CPP), and conducting Design of Experiments (DoE) and General Linear Model (GLM) analyses to establish the system's design space. The results indicate that the SVM model achieved superior fraud detection performance, with a Recall of 5.76% and an F1-Score of 0.0403, outperforming the Random Forest model, which achieved a Recall of 0.62% and an F1-Score of 0.0104. The QbD analysis identified five critical parameters, Amount, Location, Interval, Device, and Receiver Account that significantly influence fraud detection performance. Furthermore, the DoE and GLM analyses revealed that interactions among these parameters have a greater impact on the fraud rate than the individual effects of each parameter. This study establishes a design space that can serve as a foundation for suspicious transaction monitoring and support the development of a more systematic, adaptive, and operationally effective fraud detection system for the banking industry.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Mobile Banking, Fraud Detection, Quality by Design, Support Vector Machine, Mobile Banking, Fraud Detection, Support Vector Machine, Quality by Design
Subjects:	T Technology > T Technology (General) T Technology > T Technology (General) > T57.8 Nonlinear programming. Support vector machine. Wavelets. Hidden Markov models.
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26201-(S1) Undergraduate Thesis
Depositing User:	Amanda Na`ilah Neysa
Date Deposited:	31 Jul 2026 02:38
Last Modified:	31 Jul 2026 02:38
URI:	http://repository.its.ac.id/id/eprint/141535

Actions (login required)

View Item