cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Muh. Isbar Pratama
Contact Email
isbarpratama@unm.ac.id
Phone
+6285399692435
Journal Mail Official
jmathcos@unm.ac.id
Editorial Address
Kampus Parangtambung UNM, Jl. Dg. Tata Raya Prodi Matematika Lt. 3 Gd FG Jurusan Matematika FMIPA
Location
Kota makassar,
Sulawesi selatan
INDONESIA
Journal of Mathematics, Computation and Statistics (JMATHCOS)
Published by Universitas Negeri Makassar
ISSN : 24769487     EISSN : 27210863     DOI : https://doi.org/10.35580/jmathcos
Core Subject : Education,
Mathematics
Fokus yang didasarkan tidak hanya untuk penelitian dan juga teori-teori pengetahuan yang tidak menerbitkan plagiarism. Ruang lingkup jurnal ini adalah teori matematika, matematika terapan, program perhitungan, perhitungan matematika, statistik, dan statistik matematika.
Arjuna Subject : Matematika - Matematika Terapan
Articles 214 Documents
 18   19   20   21   22 
Backtesting of the Value-at-Risk Based on GARCH Model (VaR-GARCH) in Measuring Stock Market Risk Lailatul Maziyah Wildan Mufaridho; Khaerun Nisa SH; Paiz Jalaludin; Muhammad Isbar Pratama
Journal of Mathematics, Computations and Statistics Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/eyqd4722

Abstract

Accurate market risk measurement is a crucial aspect of stock portfolio management, particularly in volatile market conditions. One commonly used method for measuring market risk is Value-at-Risk (VaR). However, the conventional VaR approach often fails to capture the dynamics of volatile volatility. Therefore, this study aims to measure stock market risk using a GARCH-based Value-at-Risk approach and test the model's reliability using the Kupiec Proportion of Failures Test. The data used are daily stock price data processed into logarithmic returns. Return volatility is estimated using the GARCH(1,1) model, and the VaR value is calculated based on conditional volatility at a 5 percent significance level. VaR backtesting is then performed to identify violations and evaluate the model's validity using the Kupiec Test. The results of the study show that out of 653 observations, there were 27 VaR violations, with a Kupiec statistic value of 1.0909 and a p-value of 0.2963. A p-value greater than the significance level indicates that the VaR–GARCH model is statistically valid and able to measure market risk well. This study concludes that the VaR–GARCH approach is a reliable method in measuring stock market risk and can be used as a supporting tool in investment decision-making and risk management.
Implementation of Recurrent Neural Network with Long Short Term Memory Algorithm for Dengue Fever Prediction in Medan City Rivani Kabrina Br Surbakti; Rima Aprilia; R Maisaroh Rezyekiyah Siregar
Journal of Mathematics, Computations and Statistics Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/fcqd6q15

Abstract

. Dengue Hemorrhagic Fever (DHF) remains one of the major public health problems in Medan City due to the high incidence rate each year. Accurate prediction of DHF cases is essential as an early warning and to support health policy planning. This study aims to implement the Recurrent Neural Network (RNN) with the Long Short-Term Memory (LSTM) algorithm to predict the number of DHF cases in Medan City. The data used consist of monthly DHF cases from each public health center (puskesmas) in Medan City from January 2020 to December 2024, obtained from the Medan City Health Office. The data were preprocessed through normalization and divided into training and testing sets. The LSTM model was developed with several testing scenarios of units, epochs, and batch size, and evaluated using Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). The results showed that the LSTM model could predict DHF cases with relatively low error rates, achieving an RMSE of 2.02 and an MAE of 1.64 at the best configuration. Therefore, it can be concluded that the LSTM algorithm is effective in predicting the number of DHF cases in Medan City and can serve as a reference in prevention and disease control strategies. Keywords: Dengue Hemorrhagic Fever (DHF), Prediction, Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Time Series.
Geographically Weighted Regression Model in the Case of Unemployment in North Sumatra Muhammad Arfie Munawar; Fibri Rahkmawati; Rini Halila Nasution
Journal of Mathematics, Computations and Statistics Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/3vd63s64

Abstract

Unemployment is a major and complex issue that affects many aspects of society, particularly in regions such as North Sumatra. This issue is not merely about numbers it also concerns the welfare of the population. Each district or city exhibits varying levels of unemployment; some areas have high rates, while others are relatively low. These variations reflect a clear spatial heterogeneity. To address the significant spatial variation in the factors contributing to unemployment, this study applies the Geographically Weighted Regression (GWR) model to analyze and model unemployment in North Sumatra, taking into account the spatial variability of each independent variable’s influence. GWR is a regression method that allows model parameters to vary across geographic locations, making it possible to capture non-uniform relationships at different spatial points. The methodology involves four weighting functions adaptive Gaussian, adaptive bisquare, fixed Gaussian, and fixed bisquare to identify the most optimal model. The best-performing GWR model is then constructed, and the spatial distribution patterns of unemployment are analyzed. The data used in this study are sourced from official statistics. The results show that the adaptive bisquare GWR model provides the best performance, yielding the lowest AIC value of 130.066. Variables such as population density and population growth rate are significant in most regions. However, number of industries is only significant in certain areas, while total population and minimum wage are not significant. These findings indicate that the factors driving unemployment and their spatial distribution vary across regions, highlighting the importance of considering spatial heterogeneity.
A Comparative Performance Evaluation of Copilot, Gemini, and DeepSeek in Understanding Knot Semantic Logic Khaerati, Khaerati; Ja'faruddin, Ja'faruddin; Fadilah, Nur; Aslindawati, Nur; Fadiyah, Wulan Nuf; Ihsan, Muh.
Journal of Mathematics, Computations and Statistics Vol. 8 No. 2 (2025): Volume 08 Nomor 02 (Oktober 2025)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/7x4cwq80

Abstract

This study examines the ability of three artificial intelligence (AI) models Copilot 3.7 Sonnet, Gemini 1.5, and DeepSeek-R1 to interpret Knot Semantic Logic (KSL), a topological framework emphasizing semantic symmetry in textual structures. A qualitative descriptive design with comparative analysis was employed. Structured prompts tested the models across three stages: basic concept mastery, analysis of symmetrical sentences, and generalization to new inputs. Performance was assessed using five indicators Conceptual Accuracy, Structural Accuracy, Generalization, Consistency, and Narrative Clarity combined into a KSL-AI Index. The results show distinct performance profiles. Copilot produced accessible explanations but lacked structural precision. Gemini demonstrated stability in recognizing semantic symmetry, supported by large-scale multimodal and multilingual training, although its technical style limited accessibility. DeepSeek showed strength in detecting simple patterns and basic logic but was less consistent and struggled with complex generalization tasks. The study validates KSL as an innovative evaluation tool, extending AI assessment beyond narrative fluency to structural semantic reasoning. It concludes that Copilot is best suited for pedagogical use, Gemini for consistent analytical tasks, and DeepSeek for exploratory analysis. Future work should integrate quantitative measures, multimodal testing, and broader model comparisons.
Formulation of the Black–Scholes Model for Agricultural Insurance Premium Determination in Response to Climate Change Risk Kalfin Kalfin; Ni Luh Sri Diantini; Hisyam Ihsan
Journal of Mathematics, Computations and Statistics Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/khyhbb11

Abstract

Climate change has intensified the risk of crop failure and economic losses in the shallot farming sector, necessitating the development of insurance premium models that are adaptive to climate dynamics. This study aims to evaluate existing agricultural insurance premium models for shallots and to develop a new approach that is more responsive, equitable, and sustainable through the integration of data science and statistics. The main innovation of this research lies in the application of a modified Black–Scholes model, incorporating two key variables: rainfall index and shallot production risk. This model serves as the basis for determining insurance premiums in Tasikmalaya Regency. Simulations were conducted using primary data collected through questionnaires distributed to shallot farmers, as well as secondary data on production records and rainfall from 2016 to 2024. The estimation results reveal a positive and significant relationship between rainfall percentiles and insurance premiums, where higher rainfall levels tend to be associated with increased production and insured value, leading to higher premium rates. These findings offer practical insights for insurance companies in designing more accurate and sustainable index-based premium schemes for shallot commodities.
Clustering of Central Java Districts Based on Educational Indicators: A Comparison of K-Means and Hierarchical Methods Muhammad Syafiq; Nabila Fida Millati; Muh Akbar Idris; Anwar Fitrianto; Kevin Alifviansyah; Erfiani Erfiani
Journal of Mathematics, Computations and Statistics Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/xen35m31

Abstract

This study aims to cluster districts and municipalities in Central Java based on educational indicators and to compare the clustering performance of K-Means and Hierarchical methods. The analysis uses secondary data from the Statistical Publication of Education in Central Java Province 2024, covering eight indicators related to educational facilities, participation, and attainment. The data were standardized, explored using descriptive statistics, and analyzed using K-Means and Hierarchical clustering methods. The evaluation results show that both methods produced broadly comparable clustering structures. However, Hierarchical Clustering demonstrated slightly stronger performance in terms of cluster separation and compactness, with a higher Silhouette Index (0,591) and Dunn Index (0,320) and a lower Davies–Bouldin Index (0,501) compared with K-Means (SI 0,584, Dunn 0,225, DBI 0,562). Meanwhile, K-Means produced a more balanced partition and a higher Calinski–Harabasz Index (48,63) than Hierarchical Clustering (44,30). The clustering results reveal a clear pattern of educational disparities across the region. A small group consisting of Sukoharjo Regency and the cities of Semarang, Surakarta, Salatiga, and Magelang forms a higher-performing cluster characterized by stronger educational indicators, while most rural districts belong to a lower-performing group. These findings indicate that educational disparities in Central Java remain spatially concentrated and highlight the need for targeted policies to strengthen educational investment and improve progression to higher levels of education in less developed districts.
Comparison of K-Means and K-Medoids Methods in Grouping Provinces in Indonesia Based on Economic Development Indicators Widya Saputri Agustin; Rinda Armadhan; Sekti Kartika Dini; Handani Murda
Journal of Mathematics, Computations and Statistics Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/jmathcosv9n19866

Abstract

Economic growth in Indonesia varies significantly between provinces, reflecting disparities in welfare indicators such as poverty levels, education, and access to infrastructure. Understanding these disparities is crucial for formulating effective development policies. This study aims to cluster provinces in Indonesia based on economic development indicators 2023, with the dataset sourced from Badan Pusat Statistik (BPS). The research employs K-Means and K-Medoids clustering methods, with the optimal number of clusters determined using the Silhouette method. K-Means produced six clusters, while K-Medoids identified eight clusters. Performance evaluations using the Dunn Index (DI), Davies-Bouldin Index (DBI), and Xie-Beni Index (XBI) revealed that K-Means outperformed K-Medoids, achieving a higher DI (0.31) and lower XBI (1.78). These results indicate that K-Means with six clusters provides better separation and higher intra-cluster density compared to K-Medoids. Profiling of the clusters revealed substantial regional disparities, with some clusters exhibiting high welfare levels and others facing significant challenges in poverty, unemployment, and health issues. Cluster 1 has moderate income and development but high unemployment and health issues. Cluster 2 shows strong development and low poverty but unresolved crime. Cluster 3 has low income, minimal poverty, and health complaints. Cluster 4 excels in income and labor but struggles with poverty and crime. Cluster 5 is prosperous but faces health issues. Cluster 6 has low income, moderate poverty, and significant health challenges. This study aims to assist policymakers in designing tailored strategies to address specific weaknesses and capitalize on regional strengths to reduce disparities and enhance equitable development.
Analysis of Household Risk Factors Associated with Food Anxiety Using Boosting-Based Machine Learning Methods Nisa Nur Aisyah; Rupmana Br Butar; Mega Ramatika Putri; Lisa Amelia; Bagus Sartono; Aulia Rizki Firdawanti
Journal of Mathematics, Computations and Statistics Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/nz9epj83

Abstract

Food anxiety represents an early psychological indicator of household food insecurity and is influenced by economic vulnerability, household characteristics, and unstable access to food. West Java, as Indonesia’s most populous province, faces substantial socio-economic disparities that heighten the risk of food insecurity. Using SUSENAS 2024 data, this study aims to classify household food anxiety and evaluate the predictive performance of three boosting algorithms XGBoost, LightGBM, and CatBoost. The dataset exhibits a strong class imbalance, with only 19.1% of households categorized as food anxious, prompting the application of SMOTE and Winsorization during preprocessing. SMOTE considerably improved model performance, particularly in balanced accuracy. For XGBoost, balanced accuracy increased sharply from 0.5199 to 0.8738, while LightGBM experienced a similar improvement from 0.5261 to 0.8736. Winsorization produced only marginal additional effects. Across all scenarios, XGBoost demonstrated the highest overall performance, followed closely by LightGBM, whereas CatBoost showed limited ability to detect minority-class households. These findings underscore the effectiveness of boosting algorithms especially XGBoost enhanced by SMOTE in identifying food-anxious households and supporting data-driven, targeted food security interventions in West Java.
Comparison of OPTICS and HDBSCAN Performance in Clustering Population Administration Document Ownership in Bone Bolango Regency Adisti Dayo; Djihad Wungguli; Muhammad Rezky Friesta Payu
Journal of Mathematics, Computations and Statistics Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/s6m22g74

Abstract

Population administration is essential for public service delivery and development planning; however, disparities in population document ownership across villages remain a challenge in Bone Bolango Regency. The heterogeneous nature of the data, the presence of outliers, and variations in density patterns limit the effectiveness of classical statistical approaches in capturing the underlying distribution. Therefore, this study aims to compare two density-based clustering algorithms, Ordering Points to Identify the Clustering Structure (OPTICS) and Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN), in grouping villages based on population document ownership levels. The data were obtained from the Department of Population and Civil Registration of Bone Bolango Regency in 2024 and consist of ownership records of birth certificates, identity cards, and family cards from 165 villages. Both algorithms successfully formed two main clusters representing villages with relatively high and low levels of population document ownership. Internal validation results indicate that OPTICS outperformed HDBSCAN, achieving a Silhouette Coefficient of 0.827, a Davies–Bouldin Index of 0.242, and a Calinski–Harabasz Index of 1217.425, compared to 0.787, 1.210, and 767.866, respectively, for HDBSCAN. In conclusion, OPTICS demonstrates superior capability in producing a more coherent clustering structure for population document ownership data. Therefore, the clustering results obtained using OPTICS can serve as a supporting basis for formulating policies to promote equitable population administration services.
Latent Segmentation of Technology Adoption Behaviour Using FIMIX-PLS Awit Marwati Sakinah; Lina Listiani
Journal of Mathematics, Computations and Statistics Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026)
Publisher : Jurusan Matematika FMIPA UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35580/t5c36z45

Abstract

Digital transformation in Micro, Small, and Medium Enterprises (MSMEs) requires analytical approaches capable of capturing heterogeneous technology adoption behaviour, as homogeneous models often overlook hidden structural differences. This study investigates latent heterogeneity in MSME digital adoption using the Finite Mixture Partial Least Squares (FIMIX-PLS) approach within the Technology Acceptance Model framework based on survey data from 210 MSMEs in West Java, Indonesia. The optimal segmentation solution was determined through a combination of statistical criteria, including AIC, BIC, CAIC, and entropy values, together with practical considerations of segment size proportions to ensure parameter estimation stability and maintain generalisability given the moderate sample size. The results identify three statistically distinct latent segments with entropy values above 0.7, indicating reliable class separation. The three-segment solution produces a more balanced distribution than four-segment and five-segment. The results indicate substantial behavioral heterogeneity across user segments. Segments 1 and 2 exhibit strong positive relationships across all structural paths, with Perceived Ease of Use emerging as the dominant factor shaping attitude, which subsequently drives behavioral intention and actual use, showing stronger effects than the global model. In contrast, Segment 3 demonstrates an opposing pattern with mostly negative coefficients, suggesting low technology acceptance and potential resistance toward the system. These findings confirm that technology adoption mechanisms are not homogeneous and vary significantly across user groups

Page 21 of 22 | Total Record : 214

 18   19   20   21   22 

Filter by Year

2018 2026


Reset
Filter By Issues
All Issue Vol. 9 No. 1 (2026): Volume 09 Issue 01 (March 2026) Vol. 8 No. 2 (2025): Volume 08 Nomor 02 (Oktober 2025) Vol. 8 No. 1 (2025): Volume 08 Nomor 01 (April 2025) Vol. 7 No. 2 (2024): Volume 07 Nomor 02 (Oktober 2024) Vol. 7 No. 1 (2024): Volume 07 Nomor 01 (April 2024) Vol. 6 No. 2 (2023): Volume 06 Nomor 02 (Oktober 2023) Vol. 6 No. 1 (2023): Volume 06 Nomor 01 (April 2023) Vol. 5 No. 2 (2022): Volume 05 Nomor 02 (Oktober 2022) Vol. 5 No. 1 (2022): Volume 05 Nomor 01 (April 2022) Vol. 4 No. 2 (2021): Volume 04 Nomor 02 (Oktober 2021) Vol. 4 No. 1 (2021): Volume 04 Nomor 01 (April 2021) Vol. 3 No. 2 (2020): Volume 03 Nomor 02 (Oktober 2020) Vol. 3 No. 1 (2020): Volume 03 Nomor 01 (April 2020) Vol. 2 No. 02 (2019): Volume 02 Nomor 02 (Oktober 2019) Vol. 2 No. 01 (2019): Volume 02 Nomor 01 (April 2019) Vol. 1 No. 02 (2018): Volume 01 Nomor 02 (Oktober 2018) Vol. 1 No. 01 (2018): Volume 01 Nomor 01 (April 2018) More Issue