Penerapan Graph Neural Network dan Decision Tree untuk Sistem Rekomendasi Pinjaman Perusahaan Berdasarkan Analisis Risiko Kredit

Auliana, Fadhila (2024) Penerapan Graph Neural Network dan Decision Tree untuk Sistem Rekomendasi Pinjaman Perusahaan Berdasarkan Analisis Risiko Kredit. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06111740000047-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (4MB) | Request a copy

Abstract

Pembiayaan usaha merupakan salah satu aspek penting dalam menjalankan bisnis, terlebih lagi perusahaan kecil dan menengah. Kebutuhan dana sebagai modal usaha, kegiatan operasional harian dan pengembangan perusahaan sangatlah dibutuhkna. Salah satu sumber dana tersebut adalah pinjaman (kredit) perusahaan. Bank sebagai salah satu lembaga pembiayaan, menyediakan layanan pinjaman kredit usaha bagi perusahaan. Risiko kredit adalah isu terpenting dalam keuntungan dan pengambilan keputusan industri perbankan. Dalam pemberian kredit, pihak bank perlu melakukan analisis terhadap kondisi perusahaan sebagai upaya mengatasi risiko kredit. Oleh karena itu dalam penelitian ini dilakukan analisa menggunakan machine learning untuk menilai kelayakan pinjaman dan risiko kredit pada debitur (perusahaan). Penelitian ini dilakukan dengan metode Graph Neural Network dan Decision Tree dalam mendukung pembuatan sebuah sistem rekomendasi berbasis Content Based Filtering terhadap kelayakan pinjaman berdasarkan analisis risiko kredit. Graph Neural Network digunakan untuk menggambarkan relasi antara atribut-atribut data debitur (perusahaan). Sedangkan Decision Tree untuk mengklasifikasi atribut data debitur untuk menilai kelayakan pinjaman bagi debitur (perusahaan) serta mengetahui risiko kredit dari para debitur (perusahaan). Performa model diuji dan didapatkan hasil berupa nilai accuracy sebesar 0, 97525, precision sebesar 88, 17%, recall sebesar 81, 35%, dan f1 score sebesar 81, 88%. Model menunjukkan kinerja lebih baik dibanding hasil dengan Decision Tree saja. Hal ini menunjukkan bahwa model Graph Neural Network dan Decision Tree mampu bekerja dengan baik pada relasi kompleks dari data dan membutuhkan representasi dalam bentuk graf.
==============================================================================================================================
Business financing is one of the important aspects of running a business, especially for small and medium companies. The need for funds as business capital, daily operational activities, and company development is very much needed. One source of these funds is a company loan (credit). Banks as one of the financing institutions provide business credit loan services for companies. Credit risk is the most important issue in the profits and decision-making of the banking industry. In providing credit, the bank needs to analyze the condition of the company to overcome credit risk. Therefore, in this study, an analysis was carried out using machine learning to assess the feasibility of loans and credit risk to debtors (companies). This study was conducted using the Graph Neural Network and Decision Tree methods to support the creation of a recommendation system based on Content-Based Filtering for loan eligibility based on credit risk analysis. Graph Neural Network is used to describe the relationship between debtor (company) data attributes. While Decision Tree is used to classify debtor data attributes to assess the feasibility of loans for debtors (companies) and to determine the credit risk of debtors (companies). The model performance was tested and the results obtained were an accuracy value of 0.97525, a precision of 88.17%, a recall of 81.35%, and an f1 score of 81.88%. The model showed better performance than the results with Decision Tree alone. This shows that the Graph Neural Network and Decision Tree models can work well on complex data relations and require representation in the form of graphs.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	credit, graph neural network, decision tree, recommendation system, kredit, sistem rekomendasi
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics Q Science > QA Mathematics > QA166 Graph theory Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Fadhila Auliana
Date Deposited:	26 Aug 2024 07:32
Last Modified:	26 Aug 2024 07:32
URI:	http://repository.its.ac.id/id/eprint/114024

Actions (login required)

View Item