Komparasi Algoritma Klasifikasi C4.5 dan Naïve Bayes Berbasis Particle Swarm Optimization Untuk Penentuan Resiko Kredit

Achmad Rifai, Rizki Aulianita - STMIK Nusa Mandiri Jakarta

Abstract


Abstrak – Perkreditan memiliki peran penting dalam pembiayaan pembangunan infrastruktur dan perekonomian nasional sekarang ini. Begitu juga dengan koperasi yang merupakan pembiayaan berskala kecil yang digemari masyarakat menengah kebawah terutama dalam fasilitas simpan pinjam. Namun, fasilitas peminjaman memiliki risiko yang besar dalam kerugian yang harus ditanggung oleh koperasi, jadi dalam mengurangi risiko dalam peminjaman diperlukan metode dalam menjalankannya. Dalam penelitian ini dilakukan dengan menggunakan komparasi algoritma klasifikasi data mining C4.5 dan Naïve Bayes kemudian dilakukan optimasi berbasis Particle Swarm Optimization. Berdasarkan hasil pengujian bahwa nilai akurasi algoritma C4.5 sebesar 85.40% dan nilai akurasi algoritma Naïve Bayes sebesar 85,09%. Dari kedua algoritma tersebut kemudian dilakukan kombinasi dengan optimasi Particle Swarm Optimization, dengan hasil algoritma C4.5+PSO memiliki nilai tertinggi berdasarkan nilai accuracy sebesar 87.61%, AUC sebesar 0.860 dan precision sebesar 88.96% sedangkan nilai recall tertinggi diperoleh oleh algoritma Naïve Bayes+PSO sebesar 96.75%.

Kata Kunci: Kredit, Data Mining, Algoritma C4.5, Naïve Bayes, Optimasi

 

Abstract – Credit has an important role in financing infrastructure development and the current national economy. Likewise with the cooperative which is a small-scale financing that favored the middle to lower society, especially in savings and loan facilities. However, lending facilities have a great risk in losses that must be borne by the cooperative, so in reducing the risk in lending method required in running it. In this research is done by using comparation of data mining classification algorithm C4.5 and Naïve Bayes then done by Particle Swarm Optimization.Based on the test results that the accuracy of C4.5 algorithm of 85.40% and the accuracy of Naïve Bayes algorithm of 85.09%. The two algorithms are then combined with Particle Swarm Optimization optimization, where the C4.5 + PSO algorithm has the highest value based on accuracy value of 87.61%, AUC is 0.860 and precision is 88.96% while the highest recall value is obtained by Naïve Bayes + PSO algorithm 96.75%.

Key Word: Credit, Data Mining, Algoritm C4.5, Naïve Bayes, Optimization


Full Text:

PDF

References


Dawson, C. W. (2019). Projects in Computing and Information System A Student’s Guide (2nd ed.). England: Addison-Wesley.

Gorunescu, F. (2011). Data Mining Concept, Models and Techniques (12th ed.). Berlin Heidelberg: Springer Verlag.

Kusrini, & Luthfi, E. T. (2009). Algoritma Data Mining. Yogyakarta: Andi Offset.

Larose, D. T. (2006). Data Mining Methods and Models. Canada: John Wiley & Sons, Inc., Hoboken, New Jersey.

Liao, T. W. (2007). Recent Advances in Data Mining of Enterprise Data (6th ed.). Singapore: World Scientific.

Maimon, O., & Rokach, L. (2010). Data Mining and knowledge Discovery Handbook (Second). New York: Springer.

Rifai, A. (2016). Kajian Algoritma C4.5 Naive Bayes ,Neural Network Dan Svm Dalam Penentuan Kelayakan Pemberian Kredit. Jurnal Sistem Informasi Stmik Antar Bangsa, V NO 2(2), 176–182.

Saraswati, L., & Yadnyana, I. K. (2014). Pengaruh Struktur Pengendalian Intern Terhadap Kelancaran Pengembalian Kredit Pada Koperasi Simpan Pinjam Di Kota Denpasar. Jurnal Akuntansi Universitas Udayana, 7(1), 122–134.

Sawant, A. A., & Chawan, P. M. (2013). Study of Data Mining Techniques used for Financial Data Analysis, 2(3), 503–509.

Vercellis, C. (2009). Business Intelligence: Data Mining and Optimization for Decision Making (First). Italy: John Wiley & Sons Ltd.

Witten, I. H., Frank, E., & Hall, M. A. (2011). Data Mining : Practical Machine Learning and Tools (Third). Burlington: Morgan Kaufmann.

Zurada, J. (2010). Could decision trees improve the classification accuracy and interpretability of loan granting decisions? Proceedings of the Annual Hawaii International Conference on System Sciences, 1–9. https://doi.org/10.1109/HICSS.2010.124


Refbacks

  • There are currently no refbacks.