Muhammad, Nauval (2023) PEMODELAN ANALITIKA DATA MELALUI PENGKLASIFIKASIAN UNTUK PENENTUAN KOLEKTIBILITAS PEMBIAYAAN. Skripsi thesis, Politeknik Keuangan Negara STAN.
|
Text (Cover)
01.Cover_Nauval Muhammad_4132210017.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (107kB) |
|
|
Text (Abstrak)
02. Abstrak_Nauval Muhammad_4132210017.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (104kB) |
|
|
Text (Daftar Isi)
03.Daftar Isi_Nauval Muhammad_4132210017.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (280kB) |
|
|
Text (Bab I)
Bab I_Nauval Muhammad_4132210017.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (178kB) |
|
|
Text (Bab II)
Bab II_Nauval Muhammad_4132210017.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (311kB) |
|
|
Text (Bab III)
Bab III_Nauval Muhammad_4132210017.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (586kB) |
|
|
Text (Bab IV)
Bab IV_Nauval Muhammad_4132210017.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (1MB) |
|
|
Text (Bab IV)
Bab IV_Nauval Muhammad_4132210017.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (1MB) |
|
|
Text (Bab V)
Bab V_Nauval Muhammad_4132210017.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (126kB) |
|
|
Text (Daftar Pustaka)
Daftar Pustaka_Nauval Muhammad_4132210017.pdf Download (148kB) |
Abstract
Bagi lembaga pembiayaan, setiap penyaluran pembiayaan yang diberikan tentu memiliki risikonya tersendiri. Otoritas Jasa Keuangan membuat regulasi untuk memitigasi ini berdasarkan lima kualitas pembiayaan/kolektibilitasnya sebagai dasar dalam penghitungan CKPN. Dalam menentukan tingkat kolektibilitasnya variabel prospek usaha, kinerja dan kemampuan membayar diperhitungkan, namun terkadang ini membutuhkan effort berlebih. Disisi lain, data kolektibilitas atas pembiayaan sudah terkumpul bertahun-tahun. Penelitian ini bertujuan untuk mencari model terbaik untuk pengklasifikasian kolektibilitas pembiayaan menggunakan teknik data mining berdasarkan data historis pembiayaan selama 2014-2022. Metode penelitian yang digunakan dalam penelitian ini adalah mixed method menggabungkan penelahaan proses bisnis secara kualitatif dan analisis data numerik atas data sekunder secara kuantitatif menggunakan metode CRISP-DM. Proses business understanding dan data understanding atas 2 dataset dilakukan terhadap lebih dari 70 kolom dan tidak kurang dari 600.000 baris. Berbagai teknik dilakukan pada tahap data preparation untuk menyiapkan dataset agar siap dilakukan pemodelan. Setelah itu pemodelan dilakukan menggunakan algoritma sederhana dan yang lebih canggih. Kemudian dilakukan evaluasi performa terbaik atas kelima model. Hasilnya adalah Random Forest, Decision Trees dan k-NN memiliki performa tertinggi, dengan Decision Trees yang memiliki kecepatan komputasi paling singkat diantara ketiganya. Model ini dapat diimplementasikan untuk memberikan insight bagi LPEI untuk memitigasi risiko pembiayaan dengan lebih cepat. For financial institutions, every financing distributed has their own risks. Financial Services Authority (OJK) created regulation to mitigate them by dividing into five financing quality/ collectibility as the basis to compute allowance for doubtful account (CKPN). In order to determine collectibility rank, many variables such as business prospect, performance and ability to pay is considered, yet sometimes it requires much more effort. Meanwhile, collectibility data of financing has been collected from many years behind. This research aims to find the best model to classify financing collectibility using data mining technique based on historical financing data from 2014-2022. Research method used is mixed method that combined qualitative analysis of business processes and quantitative analysis of numerical data on secondary data using the CRISP-DM method. Business and data understanding process is done to 2 sets of data with more than 70 columns and no less than 600.000 rows. So many techniques are applied in data preparation step to make sure the dataset is ready for modelling. After that modelling is executed through simple and advanced algorithms. The next step is best performance evaluation amongst the five model. The results are that Random Forest, Decision Trees and kNN has outperform the others, while Decision Trees has the fastest compute time. This model could be implemented to give insights for LPEI to mitigate their financing risk quicker.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Uncontrolled Keywords: | kolektibilitas, kualitas pembiayaan, klasifikasi, data mining, decision trees , collectability, financing quality, classification, data mining, decision trees |
| Subjects: | PKN STAN Subject Area > Sistem Informasi |
| Divisions: | 62303 Diploma IV Akuntansi Sektor Publik |
| Depositing User: | Perpustakaan PKN STAN |
| Date Deposited: | 07 Jan 2026 01:28 |
| Last Modified: | 07 Jan 2026 01:28 |
| URI: | http://eprints.pknstan.ac.id/id/eprint/3135 |
Actions (login required)
![]() |
View Item |

