cover
Contact Name
Nusar Hajarisman
Contact Email
statistika@unisba.ac.id
Phone
+628124569091
Journal Mail Official
statistika@unisba.ac.id
Editorial Address
Program Studi Statistika Universitas Islam Bandung Jl. Ranggagading No. 8 Bandung 40116 Phone: 081572198581 Email: nusarhajarisman@unisba.ac.id
Location
Kota bandung,
Jawa barat
INDONESIA
STATISTIKA
Core Subject : Science, Education,
STATISTIKA published by Department of Statistics, Faculty of Mathematics and Natural Sciences, Bandung Islamic University as pouring media and discussion of scientific papers in the field of statistical science and its applications, both in the form of research results, discussion of theory, methodology, computing, and review books. Published biannually in May and November each.
Arjuna Subject : -
Articles 91 Documents
Analysis of Variables that Influence Digital Entrepreneurship in the Western Region of Indonesia Using Panel Data Regression Erina Herwindalita
Statistika Vol. 25 No. 1 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i1.6287

Abstract

Abstract. Technological developments and digitalization provide a new space to support various entrepreneurial and productivity activities. However, digital-based entrepreneurship in Indonesia is not yet running optimally, especially in the western region of Indonesia, which has great potential. Thus, this research aims to analyze the variables influencing digital entrepreneurship in western Indonesia in 2020-2024 using a panel data regression model. Among the three model estimation approaches, it was found that the fixed effect model with cross-sectional weight was the most effective. The analysis results show that digital entrepreneurship in the western region of Indonesia experiences growth every year. Household internet access and the high school/equivalent net enrollment rate (NER) have a significant positive effect on digital entrepreneurship and productivity scores in the western region of Indonesia in 2020 – 2024. Meanwhile, the construction cost index and the gender empowerment index have no significant effect on digital entrepreneurship and productivity scores in the western region of Indonesia in 2020 – 2024.
Clustering Analysis for Regional Variations in Beef Production: A Comparative Study of Hierarchical and Non-Hierarchical Yunna Mentari Indah; Kuswandi, Wawan
Statistika Vol. 25 No. 1 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i1.6386

Abstract

Abstract. Cluster analysis is vital for identifying regional disparities in beef production and guiding targeted livestock development, particularly on Java Island. This study aimed to classify 119 regencies/cities in Java based on 2023 beef production data using both hierarchical (single linkage, average linkage, complete linkage, centroid linkage, Ward's method) and non-hierarchical (k-means, k-medoid, fuzzy c-means) clustering techniques. Performance evaluation through the Davies-Bouldin index and silhouette score revealed that the centroid linkage method delivered the most accurate results. The analysis identified three distinct clusters: Cluster 1 consisted of 90 regencies/cities with moderate beef production, Cluster 2 comprised 6 regencies/cities with high production, and Cluster 3 included 16 regencies/cities with low production. West Java records the highest production volume, while East Java has the largest cattle population. These results highlight regional differences in production capacity, and suggest the need for targeted policy interventions. To bridge potential gaps in beef availability and address regional imbalances, policies could focus on improving beef production in low-output areas and optimizing practices in high-production regions. This study also underscores the importance of considering production levels and livestock populations in regional development strategies. The findings are expected to contribute to efforts aimed at increasing domestic beef availability and reducing reliance on imports, thereby helping to meet the nutritional and protein needs of the population.
Analysis of Factors that Explain Customer Satisfaction and Its Relationship with Repurchase Intention at Fresh Food E-Commerce in Jabodetabek Post Pandemic Agustin, Natania; Widyaningsih, Yekti; Setiadi, Rianti
Statistika Vol. 25 No. 2 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i2.4966

Abstract

Abstract. Consumers now have the flexibility to purchase fresh food both online and offline in the post-pandemic era. In this competitive landscape, only companies that can retain their customers are likely to survive. This study investigates the factors influencing customer satisfaction and its impact on repurchase intention in the context of fresh food e-commerce. Specifically, it examines the effects of product quality, fair price, perceived ease of use, service response speed, and service convenience on customer satisfaction, and further explores the moderating role of customer experience in the relationship between satisfaction and repurchase intention. Primary data were collected using a purposive sampling technique from 160 respondents who had purchased fresh food from Sayurbox, Segari, Allofresh, or Astro within the last six months. Data was analyzed using Partial Least Square, a non-parametric statistical technique suitable for modeling latent variable relationships without distributional assumptions. The results show that product quality, fair price, service response speed, and service convenience significantly and positively influence customer satisfaction. In contrast, perceived ease of use does not have a significant effect. Customer satisfaction has a strong positive influence on repurchase intention. However, customer experience does not significantly moderate the relationship between customer satisfaction and repurchase intention. These findings highlight the importance of maintaining product and service quality to enhance customer satisfaction and encourage repeat purchases in the fresh food e-commerce sector.
A Hybrid Decision Tree and K-Means Approach for Classifying Community Happiness in Bogor Regency Anwar Fajar Rizki; Dwi Fitrianti; Sri Amaliya; Bagus Sartono; Aulia Rizki Firdawanti
Statistika Vol. 25 No. 2 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i2.5590

Abstract

Abstract. Happiness is one of the key indicators for measuring the quality of life in a community. This study aims to classify the level of happiness among residents of Bogor Regency using a hybrid approach that combines Decision Trees and K-means. The research procedure consisted of data preprocessing, clustering using K-Means to form preliminary groups, and further classification through a Decision Tree to interpret the determinants of happiness. The analysis revealed that the residents of Bogor Regency can be categorized into two groups: those who are fairly happy and those who are less happy. The hybrid model achieved its best performance with a balanced accuracy of 84%, an F1-Score of 37%, and a Kappa score of 28%. Socioeconomic factors, such as marital status, family status, occupation, and the number of cigarettes smoked, were identified as the primary determinants influencing happiness levels. The main contribution of this study lies in demonstrating the effectiveness of a hybrid Decision Tree–K-Means approach for happiness classification and providing interpretable insights that are directly useful for policymakers. These findings offer strategic implications for the local government to design more inclusive socioeconomic policies that aim to enhance happiness and overall well-being among the residents of Bogor Regency.
Analysis of Stunting Data in Indonesia Using K-Means and Self Organizing Map (SOM) Allo, Caecilia Bintang Girik; Nicea Roona Paranoan; Winda Ade Fitriya B; Bobi Frans Kuddi; Feby Seru
Statistika Vol. 25 No. 2 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i2.7778

Abstract

Abstract. Stunting is a global public health concern, including in Indonesia. The Indonesian government establishes a target for stunting prevalence reduction every year. The government is aiming for a stunting prevalence of 18% in 2025. The government certainly requires policy recommendations to achieve this target. Clustering analysis can be used to identify provinces with similar characteristics or those that still require special attention based on stunting related indicators. There are several clustering methods, including K-Means and Self-Organizing Map (SOM). This study aims to classify provinces in Indonesia based on indicators related to stunting and to compare the performance of two clustering methods. Based on the obtained data, it was found that the data contains outliers. The best clustering method can be determined using the Silhouette Coefficient (SC) and Davies Bouldin Index (DBI). The results showed that the highest SC value, 0.62, was obtained using the SOM method and the lowest DBI, 0.75, was obtained also using SOM method. Two clusters were formed using the SOM method. Cluster 1 consisted of 36 provinces in Indonesia. Cluster 2 consisted of 2 provinces, namely Highland Papua and Central Papua.
Analysis of Factors Influencing Traffic Accidents in Sidoarjo Regency Using the Geographically Weighted Regression Method Aprilianti, Inggrit Delima; Ulinnuha, Nurissaidah; Intan, Putroue Keumala
Statistika Vol. 25 No. 2 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i2.7772

Abstract

Abstract. Traffic accidents are incidents that may result in trauma, injury, disability, or even death. One of the regencies in East Java Province experiencing an annual increase in traffic accident cases is Sidoarjo Regency. Geographically Weighted Regression (GWR) is a statistical approach that analyses the relationship between independent and dependent variables, taking into account spatial variation in each region. This study applies the GWR method to identify significant factors influencing the number of traffic accidents and to classify sub-regions within Sidoarjo Regency based on those factors. This study uses variables such as accident count, population density, vehicle types, gender ratio, and geographic coordinates to capture spatial differences across Sidoarjo's districts. The results indicate that the adaptive tricube kernel in GWR is the most suitable model, achieving a coefficient of determination (R²) of 99.96%. This performance indicates that the GWR model yields a slightly better fit than the multiple linear regression model, which obtained an R² of 99.86%. The types of vehicles, specifically trucks, cars, and motorcycles, are identified as significant variables in almost all districts. In Sidoarjo Regency, the districts are classified into two clusters based on the independent variables that significantly influence traffic accidents: Cluster 1, the density–vehicle accident cluster, and Cluster 2, the vehicle-only accident cluster. This classification provides a foundation for more targeted government interventions to reduce regional traffic accidents. Policy recommendations include controlling population density and improving road infrastructure in the first cluster, while focusing on vehicle safety, monitoring goods transportation, and implementing road safety campaigns in the second cluster.
Comparison of the Claim Ratio Method and the Bornhuetter-Ferguson Chain Ladder Method in Claim Reserve Calculation Paranoan, Nicea Roona; Seru, Feby; Gultom, Ayub Sahala
Statistika Vol. 25 No. 2 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i2.7934

Abstract

Abstract. Uncertainty in life brings risks that can threaten financial stability, making the existence of insurance crucial for managing such risks. One of the key elements in insurance is the management of claim reserves, which are funds allocated to meet outstanding claim obligations. This study aims to analyze and compare two claim reserve estimation methods, namely the Claim Ratio Method and the Bornhuetter-Ferguson Method, to assess the accuracy of each. The analysis is conducted by calculating claim reserve estimates using both methods based on historical claim data. The data used in this study are simulated data obtained through random sampling using Microsoft Excel. The results show that the Claim Ratio Method produced an estimate of 204,691,130, while the Bornhuetter-Ferguson Method yielded an estimate of 211,097,953. Compared to the Claim Ratio Method, the Bornhuetter-Ferguson Method provides results that are closer to reality, as it takes into account the claim development pattern in more detail, particularly for data with high variability. The study concludes that the choice of estimation method has significant implications for the financial stability of insurance companies. More accurate reserve calculations not only strengthen the solvency and operational efficiency of insurers but also reinforce policyholder trust and confidence in insurance protection. Consequently, the adoption of more robust methods such as the Bornhuetter-Ferguson is recommended, while future research is encouraged to explore alternative or hybrid models that may further improve estimation accuracy in diverse contexts.
Addressing Autocorrelation in Aquaculture Data: A Canonical Correlation Analysis of Water Quality and Shrimp Growth Supriatin, Febriyani Eka; Rahmawati, Aulia; Dailami, Muhammad
Statistika Vol. 25 No. 2 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i2.8387

Abstract

Abstract. This study evaluates the multivariate relationship between shrimp pond water quality and vannamei shrimp growth using Canonical Correlation Analysis (CCA) with autocorrelation handling. Observational data was collected from 13 ponds (3 visits/pond; N = 39) in Banyuwangi (DOC 32–50; Nov 2022–Jan 2023), encompassing pH, temperature (Temp), dissolved oxygen (DO), Average Body Weight (ABW), and Specific Growth Rate (SGR). Diagnosis indicated residual autocorrelation, so CCA inference was made robust through cluster bootstrap per pond (B = 2000), partial CCA controlling for Visit, and within-approach (demeaning per pond), as well as being triangulated with mixed-effects models and GEE AR(1). The first canonical root was significant (ρ1 = 0.648; Wilks’ Λ = 0.541; p = 0.0015) and stable based on bootstrap (median ρ1 = 0.692; 95% CI 0.340–0.894). Canonical loading shows Temp as the main contributor on the water quality side (supporting pH, low DO), while SGR dominates on the growth side and is in the opposite direction to ABW. The redundancy index indicates a moderate level of cross-set explanation (Y from X = 13.7%; X from Y = 36.8%). These findings underscore the importance of daily temperature stability (supported by pH stability) and DO monitoring in shrimp farm management.
Integration of Extreme Value Models and Vine Copula for Urban Flood Parametric Insurance Premium Mulawarman Awaloedin
Statistika Vol. 25 No. 2 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i2.8247

Abstract

Abstract. This study integrates Extreme Value Theory (EVT) and a C-Vine Copula model to design the trigger mechanism and premium structure of a flood parametric insurance scheme in Jakarta (DKI Jakarta). Using historical flood data from 2014–2020 (84 monthly observations) with an event-based analytical unit, the model incorporates maximum rainfall (mm), flood duration (days), and the number of affected neighborhood units (RW). The Generalized Extreme Value (GEV) distribution is employed to model rainfall extremes, while the Generalized Pareto Distribution (GPD) is applied to flood duration. A C-Vine Gumbel Copula is used to capture interdependence across variables. Estimation results show Kendall’s τ of 0.58 between rainfall and affected RWs, and 0.54 between flood duration and RWs, with goodness-of-fit validation (p > 0.05). The optimal trigger — rainfall > 175 mm, flood duration > 2.5 days, and > 120 affected RWs — yields a claim probability of 3.5% ± 1.2% and a basis risk of 18%. With an insured value of IDR 10 billion and a 20% loading factor, the annual premium is determined at IDR 428.4 million. The EVT–Copula integration enhances flood risk estimation accuracy and premium efficiency, providing a replicable probabilistic framework for the development of parametric insurance products in high-risk regions.
Optimizing Train-Test Splits for LSTM and MLP Models in Bitcoin Price Forecasting Accuracy Kamisan, Nur Arina Bazilah; Lee, Muhammad Hisyam; Sulandari, Winita
Statistika Vol. 25 No. 2 (2025): Statistika
Publisher : Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/statistika.v25i2.6989

Abstract

Abstract. This study investigates the application and efficiency of two machine learning models, Long-Short Term Memory (LSTM) and Multilayer Perceptron (MLP), for cryptocurrency price forecasting, using Bitcoin as a case study. MLP is a feedforward neural network that learns patterns from independent data, while LSTM is a recurrent network that remembers past information to handle sequential or time-series data. The rapid growth and volatility of cryptocurrencies underscore the need for accurate price predictions to support investor’s and trader’s decision-making. The study aims to identify the optimal train-test splitting ratio for each machine learning model and to forecast Bitcoin prices over a 120 days. The daily Bitcoin price data is obtained from the Bitcoin website recorded from January 2018 until March 2021. Model performance was evaluated using Akaike Information Criterion (AIC), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). Experimental results demonstrate that both models exhibit strong predictive capabilities; the LSTM model consistently outperforms MLP in accuracy and reliability, achieving lower MAE, MAPE, and AIC values. These findings highlight LSTM’s effectiveness for forecasting volatile financial data and provide insights into selecting appropriate data-splitting ratios to improved model performance.

Page 9 of 10 | Total Record : 91