cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Soraya
Contact Email
jurnal.varian@stmikbumigora.ac.id
Phone
+6282339979545
Journal Mail Official
jurnal.varian@stmikbumigora.ac.id
Editorial Address
Jln. Ismail Marzuki - Cilinaya - Cakranegara - Mataram 83127
Location
Kota mataram,
Nusa tenggara barat
INDONESIA
Jurnal Varian
Published by Universitas Bumigora
ISSN : -     EISSN : 25812017     DOI : https://doi.org/10.30812/varian
Core Subject : Economy, Science, Education, Social,
Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Mathematics Social Sciences Other
Jurnal Varian adalah salah satu Jurnal Ilmiah yang terdapat di Universitas Bumigora. Jurnal ini bertujuan untuk memberikan wadah atau sarana publikasi bagi para dosen, peneliti dan praktisi baik di lingkungan internal maupun eksternal Universitas Bumigora Mataram. Jurnal ini terbit 2 (dua) kali dalam 1 tahun pada periode Genap (April) dan Ganjil (Oktober). Jurnal Varian fokus memuat publikasi pada Bidang Matematika dan Statistika.
Arjuna Subject : Matematika - Statistik dan Probabilitas
Articles 10 Documents
 1 
Search results for , issue "Vol. 9 No. 1 (2026)" : 10 Documents clear
Studi Simulasi Untuk Model Regresi Nonparametrik Dengan Fungsi Kernel Kuartik Suprianto, Esmar; Sifriyani, Sifriyani; Dani, Andrea Tri Rian
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.4949

Abstract

Nonparametric regression is a method for estimating the pattern of the relationship between predictor variables and response variables when the functional form of the regression curve is unknown. One estimator applicable to nonparametric regression is the Kernel estimator. The kernel estimator has a more flexible form, and the calculations are straightforward. The performance of the Kernel estimator is significantly affected by the Kernel function and the smoothing parameter (bandwidth). The method used in this study is the Kernel estimator, applied to a simulation study using a quartic kernel for optimal bandwidth selection via generalized cross-validation (GCV). This study aims to evaluate simulation results across various combinations of sample sizes and variances and to present a prediction plot of the Quartic Kernel function based on the simulation study. The results of this study are based on the Quartic Kernel function; larger sample sizes yield smaller Mean Squared Error (MSE) and GCV values and a larger coefficient of determination. In addition to sample size, variance is also very influential. The larger the variance, the larger the MSE and GCV values, and the smaller the coefficient of determination. The results of this study are prediction plots against the simulation studies used, showing that the Quartic Kernel function is less effective at predicting simulation study results. This is also evident from the accuracy obtained across different sample sizes and data with varying levels of variance, indicating that, in simulation studies using the quartic kernel estimator, predictive performance is poorer.
Heart Disease Classification Using ROSE and I-CHAID with Cramér’s V Bias Correction Taufiq, Annurial Fitrayah; Siswanto, Siswanto; Hadijah, Hadijah; Aprilyani, Lilis Dwi Sapta
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.5767

Abstract

Machine learning applications in healthcare are increasingly important for disease classification using categorical data. The Chi-square Automatic Interaction Detection (CHAID) method is widely used, but it often produces biased results, especially with small or imbalanced datasets. To overcome this limitation, the Improved CHAID (I-CHAID) was developed by integrating bias correction on Cramér’s V. Further performance gains on imbalanced data can be achieved by combining I-CHAID with the Random Oversampling Examples (ROSE) technique. This study aims to determine significant factors influencing heart disease and to evaluate the classification accuracy of the I-CHAID method with bias correction on Cramér’s V. The research was conducted in two stages: (1) balancing the dataset with ROSE and (2) constructing a classification tree of heart disease occurrences using I-CHAID with bias correction. The proposed I-CHAID model correctly classified 98 individuals with heart disease and 110 without heart disease out of 253 test cases. However, 30 cases were undetected (false negatives), and 15 were misclassified (false positives). Overall, the model achieved an accuracy of 84.60%, outperforming the standard CHAID method without bias correction, which reached only 71.15%. The I-CHAID method with Cramér’s V bias correction proved effective in identifying key factors associated with heart disease in Yogyakarta, including generational differences, smoking habits, and dietary patterns rich in fatty and savory foods. These findings highlight the potential of the proposed framework to support more reliable early risk identification and data-driven public health decision-making, particularly when dealing with imbalanced categorical health data.
A Comparative Study of AutoSARIMAX and Long Short-Term Memory Models for Tourist Arrival Forecasting Saptarini, Dian; Saputri, Dian Syafitri Chani; Wardhana, Helna; Martono, Galih Hendro
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.5771

Abstract

This study aims to predict the number of tourist arrivals in West Nusa Tenggara (NTB) Province using two forecasting approaches: AutoRegressive Integrated Moving Average with Exogenous Variables (AutoSARIMAX) and Long Short-Term Memory (LSTM). The dataset was obtained from the Central Bureau of Statistics (BPS) of NTB and consists of international and domestic tourist arrivals and monthly inflation rates for the period 2014–2023. The research process includes data collection, preprocessing, model construction, and result evaluation. The AutoSARIMAX model is applied to capture linear relationships with exogenous variables, while LSTM is employed to model long-term nonlinear patterns. The findings reveal that the LSTM model achieved better forecasting performance, with a Mean Absolute Percentage Error (MAPE) of 2.65%, which is lower than AutoSARIMAX with 3.25%. Nevertheless, AutoSARIMAX provides valuable interpretability regarding the influence of inflation on tourist arrivals. Overall, the comparison between the two models indicates that LSTM is more effective for time-series forecasting of tourist arrivals, while AutoSARIMAX remains useful for analyzing causal relationships. These insights can support decision-making in tourism planning, particularly in anticipating fluctuations driven by economic and external factors.
Segmentation of Teachers Using Gower-Based Hierarchical Clustering Kaloka, Tesdiq Prigel; Yudhistira, Danang Bagus; Mariyam, Mariyam; Azzahra, Hurul A’ini Sekar; Widagdo, Muhammad Rudito
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.5903

Abstract

The core issue addressed in this study is the reliance of teacher development and welfare policies on aggregated indicators that obscure variations in teachers’ demographic, professional, and economic conditions. This research aims to identify teacher profiles across multiple educational levels. The method used in this research is hierarchical clustering utilizing Gower distance applied to mixed-type survey data collected from 376 teachers across all educational levels. The analysis incorporates demographic, professional, and socioeconomic variables, including age, education, years of service, income, economic class, number of dependents, income satisfaction, and interest in technology. The analysis identifies two distinct teacher clusters. The first cluster is characterized by more experienced teachers with longer service periods, relatively stable financial conditions, and higher income satisfaction, while the second cluster comprises younger teachers with shorter teaching experience, lower income levels, and lower financial satisfaction. These findings highlight substantial heterogeneity among teachers and suggest that teacher development and welfare policies should be formulated in a differentiated manner by considering career stages and economic conditions, thereby enabling more targeted and data-driven policy interventions.  
Comparing SOM, DBSCAN, and K-Affinity Propagation in Labor Economic Patterns Nurmayanti, Wiwit Pura; Yuniarti, Desi; Siringoringo, Meiliyani; Purnamasari, Ika; Putri, Desi Febriani; Hasanah, Siti Hadijah
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.5933

Abstract

The objective of this research is to identify the most effective clustering method for grouping Indonesian provinces by labor–economic indicators to support more precise, data-driven policy formulation. Regional disparities in Indonesia’s economic growth, driven by unequal labor characteristics, remain a significant obstacle to achieving inclusive development. An analytical approach capable of grouping provinces by labor and economic indicators is therefore essential. This study applies a comparative clustering analysis using three unsupervised algorithms: Self-Organizing Maps (SOM), Density-Based Spatial Clustering of Applications with Noise (DBSCAN), and K-Affinity Propagation (K-AP). The dataset consists of five key indicators, namely economic growth, total population, labor force, employment rate, and average wage level obtained from Statistics Indonesia (BPS) for the year 2024. The clustering performance is evaluated using internal validation criteria based on the ratio of within-cluster variation (Sw) to between-cluster variation (Sb), where a smaller ratio indicates more compact, well-separated clusters. The results show that each method produces different clustering structures. SOM and DBSCAN generate three clusters with varying provincial distributions, whereas K-AP produces five clusters with more balanced, representative groupings. The evaluation results indicate ratios of 3.1906 for SOM, 0.2000 for DBSCAN, and 0.1779 for K-AP, indicating that K-AP provides the most optimal clustering performance. These findings confirm that K-Affinity Propagation is the most effective and stable method for classifying Indonesian provinces by labor and economic characteristics. The outcomes of this study provide empirical insights and analytical references for labor-driven economic policy formulation and data-driven regional development planning in Indonesia.
Principal Component Analysis and Agglomerative Hierarchical Clustering for Assessing the Condition of MSMEs Assisted by the Department of Cooperatives and MSMEs Dewi, Ardiana Fatma; Ahadiyah, Kurnia
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.6035

Abstract

Usaha Mikro, Kecil dan Menengah (UMKM) have an important role in the growth of the Indonesian economy. To achieve these hopes certainly requires a strategy. One way is to formulate policies based on information adapted to local conditions. One of the right ways to conduct this research is through data mining. There are techniques in data mining and one of the techniques that can be used is clustering with the Agglomerative Hierarchical Clustering Algorithm with Principal Component Analysis (PCA). Cluster analysis aims to group objects based on their characteristics. This research aims to determine the appropriate distribution strategy for business capital assistance. In grouping UMKM assisted by the Department of Cooperatives and UMKM of Kediri City based on several indicators measured by business capital, turnover, profits, human resources, marketing methods, government capital assistance, type of business, and place of business, it was found that the optimal algorithm used was complete linkage. With a cophenetic correlation value obtained of 0.733. Based on good internal cluster validation through silhouette values ​​based on the characteristics possessed by UMKM actors, the number of representative clusters is 3 clusters. An interesting finding is that the third cluster has not had access to government assistance programs. Based on the results of this research, it can be concluded that the allocation of government capital assistance is not fully evenly distributed and is not optimal in achieving the goal of increasing the competitiveness of UMKM.
Evaluating Random Forest Regression for Air Quality Prediction Izabi, Muh. Basyar; Annas, Suwardi; Ahmar, Ansari Saleh
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.6046

Abstract

Air pollution is a growing environmental issue in Makassar due to rapid urban development and increasing transportation activity. This study aims to model and predict air pollutant concentrations using the Random Forest (RF) regression method. The data consist of daily PM2.5, PM10, CO, NO2, SO2, and O3 measurements from September 2024 to September 2025, totaling 395 observations. Missing values (14.05%) were addressed using a hybrid approach combining linear interpolation and multiple linear regression. The RF model was trained under two data-split scenarios (70:30 and 80:20) and evaluated using SMAPE, RMSE, MAE, and R2. The results show that the 80:20 configuration provides the best predictive accuracy. CO and O3 yield the most accurate predictions with SMAPE values of 9.75% and 10.87%, and R2 of 0.973 and 0.964, respectively. PM2.5 and PM10 also show strong performance, with R2 values above 0.84. These results indicate that the RF model effectively captures pollutant variability and provides reliable forecasts. Overall, Random Forest has been shown to be a robust and accurate method for predicting air quality in Makassar, supporting environmental monitoring and early warning systems. Despite its strong performance, this study is limited to two data-partition schemes and does not incorporate temporal deep-learning architectures. Future studies may investigate hybrid ensembles or deep learning approaches to determine whether incorporating sequential modeling further enhances predictive stability.
A Bayesian Ordinal Analysis of Students’ Progression Across Van Hiele Levels Ruslau, Maria Fransina Veronica; Pratama, Rian Ade; Meirista, Etriana
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.6079

Abstract

Understanding students’ progression across hierarchical levels of geometric reasoning requires analytical approaches that respect ordinal structure and quantify uncertainty. The purpose of this study is to model students’ progression across Van Hiele levels probabilistically and to estimate level-specific transition probabilities under uncertainty using paired pretest–posttest data from Grade 8 and Grade 9 students. The method used in this study is Bayesian ordinal regression with a cumulative logit specification estimated via MCMC sampling. Van Hiele levels were modeled as ordered categorical outcomes, and posterior summaries, transition matrices, and posterior predictive checks were used to characterize progression and assess model adequacy. The results indicate that progression is predominantly upward in both grades, with negligible posterior support for regression. Grade-dependent differences are evident: Grade 8 shows broader, more heterogeneous transitions from a uniformly low baseline, whereas Grade 9 exhibits more constrained, incremental progression from a higher, more dispersed initial distribution. Posterior predictive checks confirm that the model adequately reproduces the observed posttest patterns, supporting the validity of the Bayesian ordinal specification. Pedagogically, these findings imply that students at lower baseline levels tend to undergo broader conceptual shifts, whereas those at higher levels require sustained, targeted instructional support to advance further. These findings indicate that baseline ordinal structure shapes progression dynamics and that Bayesian ordinal modeling offers a coherent alternative to significance-based approaches for analyzing hierarchical learning outcomes. Educationally, this underscores the need to align instruction with students’ initial Van Hiele levels to support optimal conceptual advancement.
Forecasting Inflation Based on Money Supply and Interest Rates Using a Transfer Function Model Rispanzira, Kirana; Mikhratunnisa, Mikhratunnisa
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.6143

Abstract

Inflation is a condition in which the prices of goods and services in a country continuously increase over an extended period. Uncontrolled inflation may lead to a decline in the value of currency, economic instability, and rising poverty levels. Several factors that influence inflation include the amount of money in circulation and Indonesia's interest rates. This study aims to model and forecast inflation in Indonesia using a multi-input transfer function model based on the amount of money in circulation and interest rates as input variables. The dataset consists of monthly observations from January 2001 to November 2025, obtained from Bank Indonesia and the Central Statistics Agency (BPS). The general stages in this study include examining data stationarity, identifying and determining the best ARIMA model for input and output series, determining the multi-input transfer function model, and forecasting. The results indicate that the transfer function model with order (0,0,0)(0,2,1)[0,0,2] provides the best performance in forecasting inflation. This model is able to capture the deflation phenomenon in early 2025 and the increasing inflation movement in the following months. The forecast for 2026 also shows a fluctuating pattern that describes macroeconomic dynamics, including changes in liquidity, interest rate policy, and global commodity conditions. A MAPE value of 1.3975% reflects excellent forecasting accuracy, indicating that the model effectively captures actual inflation dynamics. This study confirms that the transfer function approach is effective for modeling the dynamic relationship between monetary variables and inflation for forecasting inflation in Indonesia.
Forecasting the Exchange Rate of the IDR Against the USD Using the ARIMA and Exponential Smoothing Models Yostira, Lulu; Mikhratunnisa, Mikhratunnisa
Jurnal Varian Vol. 9 No. 1 (2026)
Publisher : Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/varian.v9i1.6144

Abstract

The exchange rate of the Rupiah against the US Dollar is one of the macroeconomic indicators that is volatile and affects economic stability. Therefore, a forecasting method that can produce accurate predictions is needed. This study aims to forecast the exchange rate of the Indonesian Rupiah against the US Dollar and compare the performance of the Autoregressive Integrated Moving average (ARIMA) and Exponential Smoothing models. The data used is monthly time series data on the exchange rate of the Indonesian Rupiah against the US Dollar from January 2001 to December 2025, obtained from the Ministry of Trade of the Republic of Indonesia. The stages of analysis in this study are data stationarity testing, determining the best ARIMA model based on parameter significance and assumption fulfillment (residuals are white noise and normally distributed), determining the best exponential smoothing model, forecasting, and evaluating the forecasting results.The results show that the best ARIMA model formed is ARIMA(3,1,3) with a MAPE value of 2.0624%, while the Exponential Smoothing model produces a MAPE value of 1.2687%. A comparison of MAPE values shows that the Exponential Smoothing model has a lower forecasting error rate than the ARIMA model. Therefore, in this study, the exponential smoothing model is considered more accurate and more suitable for forecasting the exchange rate of the rupiah against the US dollar during the research period.

Page 1 of 1 | Total Record : 10

 1 

Filter by Year

2026 2026


Reset
Filter By Issues
All Issue Vol. 9 No. 1 (2026) Vol. 8 No. 3 (2025) Vol. 8 No. 2 (2025) Vol 8 No 1 (2024) Vol. 8 No. 1 (2024) Vol. 7 No. 2 (2024) Vol 7 No 2 (2024) Vol 7 No 1 (2023) Vol. 7 No. 1 (2023) Vol 6 No 2 (2023) Vol 6 No 1 (2022) Vol 5 No 2 (2022) Vol 5 No 1 (2021) Vol 4 No 2 (2021) Vol 4 No 1 (2020) Vol 3 No 2 (2020) Vol 3 No 1 (2019) Vol 2 No 2 (2019) Vol 2 No 1 (2018) Vol 1 No 2 (2018) Vol 1 No 1 (2017) More Issue