Garuda - Garba Rujukan Digital

The Classification of Insurance Claim Risk Using the Multilayer Perceptron Method Handamari, Endang Wahyu; Sa'adah, Umu; Arifin, Maulana Muhamad
SAINTEKBU Vol. 17 No. 01 (2025): Vol. 17 (01) January 2025
Publisher : KH. A. Wahab Hasbullah University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32764/saintekbu.v17i01.5178

Policyholders purchase insurance policies to protect themselves or their assets from potential financial risks in the future. Insurance guarantees that if an event covered by the policy occurs, the insurance company will provide compensation according to the agreed terms. Insurance companies conduct risk assessments for each policyholder to determine the premium that must be paid, making it essential to classify risk categories accurately. The Multilayer Perceptron (MLP) is one method used for classification problems. It is a machine learning algorithm belonging to the family of artificial neural networks. MLP is a flexible algorithm that can solve various classification problems, including those with complex features and non-linear relationships between input and output variables. The result of this research is the development and implementation of a Multilayer Perceptron method to classify risk categories. The evaluation of the Multilayer Perceptron model for risk classification shows satisfactory performance. Based on the classification report from training and test data, the model does not exhibit overfitting or underfitting.

Implementasi Metode Bayesian untuk Menghitung Premi Produk Asuransi Kendaran Bermotor dengan Pendekatan Monte Carlo Markov Chain Situmorang, Boy Nathanael; A’la, Kevina Alal; Arvianti, Aurellia; Yusuf, Feby Indriana; Handamari, Endang Wahyu
Euler : Jurnal Ilmiah Matematika, Sains dan Teknologi Volume 13 Issue 2 August 2025
Publisher : Universitas Negeri Gorontalo

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37905/euler.v13i2.32930

Accurate premium determination is a fundamental aspect of risk management in motor vehicle insurance. This study implements the Bayesian method using a Markov Chain Monte Carlo (MCMC) approach to calculate the net premium. The aggregate claim model is constructed from a claim frequency distribution (Poisson) and a claim severity distribution (Generalized Extreme Value (GEV)), with the GEV distribution specifically chosen to model extreme claim risk. The analysis utilizes generated data for the period 2018–2024, with parameters derived from the historical data of PT Asuransi Jasa Indonesia Purwokerto (2013–2017). Parameter estimation, performed via OpenBUGS software, was validated to have achieved good convergence (MC-error ). Based on the estimated parameters, a premium of IDR 397.502.000 was obtained, calculated using the net premium principle from the expected value of aggregate claims. These results demonstrate that the Bayesian MCMC approach is effective for producing a robust premium estimation, contributing a pricing framework that explicitly accounts for extreme value claims.

VALUE AT RISK ESTIMATION FOR STOCK PORTFOLIO USING THE ARCHIMEDEAN COPULA APPROACH Saifullah, Mohammad Dicky; Sa'adah, Umu; Andawaningtyas, Kwardiniya; Handamari, Endang Wahyu
BAREKENG: Jurnal Ilmu Matematika dan Terapan Vol 18 No 3 (2024): BAREKENG: Journal of Mathematics and Its Application
Publisher : PATTIMURA UNIVERSITY

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30598/barekengvol18iss3pp1779-1790

Investment is one of the many ways to achieve future profits. One form of investment that is widely made is stocks. The return obtained in investing in stocks is potentially higher than other investment alternatives, but the risks borne are amplified, so it is necessary to analyze these risks that may occur. In this study, the Archimedean copula method is used to estimate the Value at Risk on shares of PT Bank Rakyat Indonesia Tbk (BBRI) and PT Telekomunikasi Indonesia Tbk (TLKM) for the period September 1, 2021, to August 31, 2023. The stock data is used to determine the Archimedean copula model and calculate the estimated value of Value at Risk (VaR) on the stock return portfolio using the Archimedean copula approach. The Archimedean copula models used are the Clayton copula model, Gumbel copula, and Frank copula. Of the three Archimedean copula models, the best model was selected by looking at the largest Maximum Likelihood Estimation (MLE) value. In this study, the log-likelihood value of Clayton copula is 7.958, Gumbel copula is 6.663, and Frank copula is 8.398. Therefore, Frank copula is the best Archimedean copula model with the largest log-likelihood value of 8.398 for the said data. Then the VaR estimation is done with the Frank copula model. The Value at Risk estimation results based on the Frank copula model show maximum loss rates of -0.0277 at the 90% confidence level, -0.0363 at the 95% confidence level, and -0.0516 at the 99% confidence level.

Improving multilayer perceptron on rainfall data using modified genetics algorithm Marji, Marji; Mahmudi, Wayan Firdaus; Handamari, Endang Wahyu; Santoso, Edy; Arifin, Maulana Muhamad
IAES International Journal of Artificial Intelligence (IJ-AI) Vol 14, No 5: October 2025
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijai.v14.i5.pp3994-4005

Rainfall prediction is essential for managing water resources, agriculture, and disaster response, particularly in regions affected by climate variability. This study introduces a modified genetic algorithm (MGA) to optimize hyperparameters of a multilayer perceptron (MLP) for rainfall forecasting. The MGA incorporates elitism to retain top-performing solutions and adaptive selection based on model accuracy. The proposed MGA–MLP model was tested on rainfall datasets from Australia and Indonesia (BMKG). Experimental results show that configurations with two hidden layers, rectified linear unit (ReLU) activation and limited-memory Broyden Fletcher Goldfarb Shannon (LBFGS) optimizer, a learning rate of 0.001 and 1000 epochs consistently delivered strong performance. The model achieved accuracies of 86.02% and 79.05%, respectively. These findings indicate that MGA significantly improves MLP performance and provides a reliable, generalizable method for rainfall prediction across diverse climatic conditions.

Usage Pattern Exploration of Effective Contraception Tool Handamari, Endang Wahyu
Journal of Research in Mathematics Trends and Technology Vol. 1 No. 1 (2019): Journal of Research in Mathematics Trends and Technology (JoRMTT)
Publisher : Talenta Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32734/jormtt.v1i1.750

Determination of methods or contraception tool used by acceptors to support the Family Planning (â€œKeluarga Berencanaâ€) is a problematic. In choosing methods or contraception tool, the acceptor must consider several factors, namely health factor, partner factor, and contraceptive method. Each method or contraception tool which is used has its advantages or disadvantages. Although it has been considering the advantages and disadvantages, it is still difficult to control fertility safely and effectively. Consequently acceptor change the method or a contraception tool that is used more than once. In order acceptors get the appropriate contraception tool then the patterns of changing in the usage of effective methods or contraception tool is determined. One of the methods that can be used to look for the patterns of changing in the usage of contraception tool is data mining. Data mining is an interesting pattern extraction of large amounts of data. A pattern is said to be interesting if the pattern is not trivial, implicit, previously unknown, and useful. The patterns presented should be easy to understand, can be applied to data that will be predicted with a certain degree, useful, and new. The early stage before applying data mining is using k nearest neighbors algorithm to determine the factors shortest distance selecting the contraception tool. The next step is applying data mining to usage changing data of method or contraception tool of family planning acceptors which is expected to dig up information related to acceptor behavior pattern in using the method or contraception tool. Furthermore, from the formed pattern, it can be used in decision making regarding the usage of effective contraception tool. The results obtained from this research is the k nearest neighbors by using the Euclidean distance can be used to determine the similarity of attributes owned by the acceptors of Family Planning to the training data is already available. Based on available training data, it can be determined the usage pattern of contraceptiion tool with the concept of data mining, where the acceptors of Family Planning are given a recommendation if the pattern is on the training data pattern. Conversely, if the pattern is none match, then the system does not provide recommendations of contraception tool which should be used.

The Classification of Insurance Claim Risk Using the Multilayer Perceptron Method Handamari, Endang Wahyu; Sa'adah, Umu; Arifin, Maulana Muhamad
SAINTEKBU Vol. 17 No. 01 (2025): Vol. 17 (01) January 2025
Publisher : KH. A. Wahab Hasbullah University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32764/saintekbu.v17i01.5178

Policyholders purchase insurance policies to protect themselves or their assets from potential financial risks in the future. Insurance guarantees that if an event covered by the policy occurs, the insurance company will provide compensation according to the agreed terms. Insurance companies conduct risk assessments for each policyholder to determine the premium that must be paid, making it essential to classify risk categories accurately. The Multilayer Perceptron (MLP) is one method used for classification problems. It is a machine learning algorithm belonging to the family of artificial neural networks. MLP is a flexible algorithm that can solve various classification problems, including those with complex features and non-linear relationships between input and output variables. The result of this research is the development and implementation of a Multilayer Perceptron method to classify risk categories. The evaluation of the Multilayer Perceptron model for risk classification shows satisfactory performance. Based on the classification report from training and test data, the model does not exhibit overfitting or underfitting.

EVALUATING NEARMISS AND SMOTE FOR VEHICLE INSURANCE FRAUD CLAIM CLASSIFICATION WITH A RANDOM FOREST CLASSIFIER Yusuf, Feby Indriana; Handamari, Endang Wahyu
VARIANCE: Journal of Statistics and Its Applications Vol 7 No 2 (2025): VARIANCE: Journal of Statistics and Its Applications
Publisher : Statistics Study Programme, Department of Mathematics, Faculty of Mathematics and Natural Sciences, University of Pattimura

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30598/variancevol7iss2page219-230

This study evaluates the detection of fraudulent car insurance claims in unbalanced data by comparing two resampling techniques, namely NearMiss (undersampling) and SMOTE (oversampling), combined with Random Forest. The public dataset, consisting of 1,000 observations and 40 features, was preprocessed for missing value handling, label encoding, and min–max normalization, and split into 70% training data and 30% test data. Three scenarios were evaluated: original data (unbalanced), NearMiss, and SMOTE, using accuracy, precision, sensitivity (recall), specificity, and F1-score evaluations. The analysis results show that NearMiss provides the most balanced performance for antifraud purposes, with a sensitivity of 0.865, an F1-score of 0.667, and an accuracy of 0.787. For the original unbalanced data, the model achieved a sensitivity of 0.297 and an accuracy of 0.767. SMOTE achieved the highest precision (0.567) and accuracy (0.783), but its sensitivity was lower than that of NearMiss. These findings confirm that the selection of resampling techniques must be aligned with operational objectives: NearMiss is more appropriate when the priority is to capture as many fraud cases as possible, while SMOTE is more suitable when false positive control is prioritized.

Title

Found 27 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 27 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 27 Documents
Search