Optimizing Support Vector Machine Parameters with Genetic Algorithm for Credit Risk Assessment

Manurung, Jonson; Mawengkang, Herman; Zamzami, Elviawaty Muisa

doi:10.1088/1742-6596/930/1/012026

Cited by 13 publications

(8 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Recently, many researches have been presented in machine learning database improvement area and data base storage techniques. The presented research in [13] and [14] prove the power of machine learning algorithms to deal with storage and retrieval using document-based or relational data storage. Other researches concentrate on enhancing data storage performance for data sets, they illustrate the necessity of distributed computing systems improvement.…”

Section: Related Workmentioning

confidence: 87%

Enhanced Data Lake Clustering Design based on K-means Algorithm

Kachaoui¹,

Belangour²

2020

IJACSA

View full text Add to dashboard Cite

In recent years, Big Data requirements have evolved. Organizations are trying more than ever to accent their efforts on industrial development of all data at their disposal and move further away from underpinning technologies. After investing around Data Lake concept, organizations must now overhaul their data architecture to face IoT (Internet of Things) and AI (Artificial Intelligence) expansion. Efficient and effective data mapping treatments could serve in understanding the importance of data being transformed and used for decisionmaking process endorsement. As current relational databases are not able to manage large amounts of data, organizations headed towards NoSQL (Not only Structured Query Language) databases. One such known NoSQL database is MongoDB, which has a high scalability. This article mainly put forward a new data model able to extract, classify, and then map data for the purpose of generating new more structured data that meet organizational needs. This can be carried out by calculating various metadata attributes weights, which are considered as important information. It also processed on data clustering stored into MongoDB. This categorization based on data mining clustering algorithm named K-Means.

show abstract

Section: Related Workmentioning

confidence: 87%

Enhanced Data Lake Clustering Design based on K-means Algorithm

Kachaoui¹,

Belangour²

2020

IJACSA

View full text Add to dashboard Cite

show abstract

“…Algoritme klasifikasi C4.5 dan Naïve Bayes dapat digunakan untuk membantu pengambilan keputusan pemberian kredit dan menganalisa kemampuan nasabah dalam membayar kredit [6]. Algoritme genetika diusulkan dalam [7] untuk diterapkan sebagai algoritme pencarian nilai parameter optimal sehingga meningkatkan akurasi klasifikasi terbaik pada Support Vector Machine (SVM). Algoritma information gain digunakan dalam [8] untuk menghitung bobot dari masing masing atribut dataset persetujuan kredit dengan klasifikasi menggunakan algoritme K-Nearest Neighbour (KNN).…”

Section: Pendahuluanunclassified

“…mewakili fitur yang ada di Tabel 1. 3: Dengan subtitusi X, rule yang didapat selanjutnya untuk dimasukkan nilai dataset Penelitian ini menyertakan atribut individu nasabah seperti status (menikah, lajang, janda/duda), status tempat tinggal, jumlah tanggungan, pendidikan yang tidak terdapat pada [6], [7], dan [9], sehingga menghasilkan akurasi yang lebih tinggi. Penerapan algoritme Naïve Bayes menghasilkan akurasi lebih tinggi dari [9] yang menerapkan forward selection pada algoritme Naïve Bayes dengan hasil akurasi sebesar 71,97%.…”

Section: Metode Penelitianunclassified

“…Penerapan algoritme decision tree menghasilkan akurasi yang lebih tinggi dari hasil klasifikasi C4.5 [6] yang menghasilkan akurasi sebesar 88,90%. Penerapan algoritme SVM pada penelitian ini mengasilkan akurasi yang lebih tinggi dibanding [7] yang menerapkan algoritme genetika sebagai algoritme pencarian nilai parameter optimal pada SVM dengan hasil akurasi kernel linear 85,12%, polinomial 81,76%, RBF 77,22%, dan Sigmoid 78,70%.…”

Section: Metode Penelitianunclassified

See 1 more Smart Citation

Algoritma Naïve Bayes, Decision Tree, dan SVM untuk Klasifikasi Persetujuan Pembiayaan Nasabah Koperasi Syariah

Nurajijah¹,

Riana²

2019

Jurnal Teknologi dan Sistem Komputer

View full text Add to dashboard Cite

The decision on financing approval in sharia cooperatives has a high risk of the inability of customers to pay their credit obligations at maturity or referred to as bad credit. To maintain and minimize risk, an accurate method is needed to determine the financing agreement. The purpose of this study is to classify sharia cooperative loan history data using the Naïve Bayes algorithm, Decision Tree and SVM to predict the credibility of future customers. The results showed the accuracy of Naïve Bayes algorithm 77.29%, Decision Tree 89.02% and the highest Support Vector Machine (SVM) 89.86%.

show abstract

“…Genetic Algorithm (GA) is relatively complicated but can locate global optimum while Particle Swarm Optimization (PSO) is not complex and possesses high convergence rate. Both of these methods are continually applied to parameters optimization of Gaussian kernel SVM [4,9,25,30,33,37] .…”

Section: Introductionmentioning

confidence: 99%

Parameter selection of Gaussian kernel SVM based on local density of training set

Yang

Peng

et al. 2020

Inverse Problems in Science and Engineering

View full text Add to dashboard Cite

Support vector machine (SVM) is regarded as one of the most effective techniques for supervised learning, while the Gaussian kernel SVM is widely utilized due to its excellent performance capabilities. To ensure high performance of models, hyperparameters, i.e. kernel width and penalty factor must be determined appropriately. This paper studies the influence of hyperparameters on the Gaussian kernel SVM when such hyperparameters attain an extreme value (0 or ∞ ). In order to improve computing efficiency, a parameter optimization method based on the local density and accuracy of Leave-One-Out method are proposed. Kernel width of each sample is determined based on the local density needed to ensure a higher separability in feature space while the penalty parameter is determined by an improved grid search using the Leave-One-Out method.A comparison with grid method is conducted to verify validity of the proposed method. The classification accuracy of five real-life datasets from UCI database are 0.9733, 0.9933, 0.7270, 0.6101 and 0.8867, which is slightly superior to the grid method. The results also demonstrate that this proposed method is computationally cheaper by 1 order of magnitude when compared to the grid method.

show abstract

Optimizing Support Vector Machine Parameters with Genetic Algorithm for Credit Risk Assessment

Cited by 13 publications

References 4 publications

Enhanced Data Lake Clustering Design based on K-means Algorithm

Enhanced Data Lake Clustering Design based on K-means Algorithm

Algoritma Naïve Bayes, Decision Tree, dan SVM untuk Klasifikasi Persetujuan Pembiayaan Nasabah Koperasi Syariah

Parameter selection of Gaussian kernel SVM based on local density of training set

Contact Info

Product

Resources

About