Article Per Year (5 Year)
p-Index From 2021 - 2026
2.307
P-Index
This Author published in this journals
All Journal Techno.Com: Jurnal Teknologi Informasi Zero : Jurnal Sains, Matematika, dan Terapan Jambura Journal of Mathematics TELKA - Telekomunikasi, Elektronika, Komputasi dan Kontrol Inferensi Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistika Euler : Jurnal Ilmiah Matematika, Sains dan Teknologi The Indonesian Journal of Computer Science Journal of Mathematics, Computation and Statistics (JMATHCOS) Indonesian Journal of Statistics and Its Applications
Aulia Rizki Firdawanti
Department Of Statistics, IPB University
Agriculture, Biological Sciences & Forestry Chemistry Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Education Electrical & Electronics Engineering Engineering Mathematics Physics Social Sciences Other

Published : 15 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 15 Documents
Search

 1   2 
ANALISIS KINERJA MODEL STACKING BERBASIS RANDOM FOREST DAN SVM DALAM KLASIFIKASI RUMAH TANGGA BERDASARKAN GARIS KEMISKINAN MAKANAN DI PROVINSI JAWA BARAT Ghiffary, Ghardapaty Ghaly; Amanda, Nabila Tri; Ardhani, Rizky; Sartono, Bagus; Firdawanti, Aulia Rizki
Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistika Vol. 5 No. 3 (2024): Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistik
Publisher : LPPM Universitas Bina Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46306/lb.v5i3.856

Abstract

The stacking method is an ensemble technique in machine learning that combines predictions from several base models to improve classification accuracy. This research applies the stacking method with two machine learning algorithms, namely Random Forest and Support Vector Machine (SVM) as base learners and logistic regression as a meta learner. This study aims to develop a classification model to identify households based on the food poverty line in West Java Province. The data used is KOR and household data in West Java Province sourced from the 2023 BPS National Socio-Economic Survey (Susenas). The variables used consisted of 24 independent variables with food poverty level as the response variable. Modeling was conducted using feature selection using Recursive Feature Elimination (RFE) and class imbalance handling using the ADASYN method. The results showed that the stacking model was superior to the single model with a balance accuracy of 0.81, sensitivity of 0.72, and specificity of 0.89. Feature importance analysis identified that calorie consumption, expenditure on cigarettes, meat and fruits, and expenditure on rice, eggs and other commodities contributed the most to the classification households based on the food poverty line in West Java Province.
STUDI KOMPARASI METODE SVM-SMOTE DAN SMOTE-TOMEK DALAM MENGATASI IMBALANCE CLASS MENGGUNAKAN MODEL XGBOOST PADA KLASIFIKASI RUMAH TANGGA PENERIMA KUR Yanuari, Eka Dicky Darmawan; Yudhianto, Rachmat Bintang; Ulfia, Ratu Risha; Sartono, Bagus; Firdawanti, Aulia Rizki
Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistika Vol. 5 No. 3 (2024): Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistik
Publisher : LPPM Universitas Bina Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46306/lb.v5i3.857

Abstract

This study aims to compare the SMOTE, SVM-SMOTE, and SMOTE-Tomek methods using the XGBoost model in overcoming the problem of class imbalance and to determine the factors that affect the status of KUR recipients in West Java Province. Three XGBoost models with class balancing techniques SMOTE, SVM-SMOTE and SMOTE-Tomek were applied to SUSENAS data of West Java Province in 2023 consisting of 1 response variable and 19 predictor variables. The results showed that the XGBoost model with the SMOTE balancing method produced better accuracy in overall data classification, but was less effective in classifying minority classes as reflected by low sensitivity and F1-Score values. The XGBoost model with the SMOTE-Tomek balancing method showed better performance in capturing minority classes with higher sensitivity and F1-Score values. The most influential variables in this model in order are per capita expenditure, urban/rural classification, motorcycle ownership, dwelling wall materials and land ownership. Per capita expenditure has the largest influence on the classification of KUR recipients, indicating that household financial management is a major factor in lending decisions. Urban/rural classification and motorcycle ownership also contributed significantly, reflecting differences in social and economic access between regions. Overall, economic factors, infrastructure and social accessibility are the main considerations in determining KUR recipient households in West Java Province.
A Analisis Perbandingan Kinerja Metode Ensemble Bagging dan Boosting pada Klasifikasi Bantuan Subsidi Listrik di Kabupaten/Kota Bogor Cintari, Nanda Putri; Alifviansyah, Kevin; Tsabitah, Dhiya Ulayya; Sartono, Bagus; Firdawanti, Aulia Rizki
The Indonesian Journal of Computer Science Vol. 13 No. 6 (2024): The Indonesian Journal of Computer Science
Publisher : AI Society & STMIK Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33022/ijcs.v13i6.4537

Abstract

The classification of electricity subsidy recipients is an crucial step to ensure that the government's social assistance program is distributed in a targeted manner, so an appropriate analysis method is needed. This research compares the Bagging and Boosting ensemble methods for the classification of households receiving electricity subsidies in Bogor Regency and City using Susenas 2023 data totaling 2002 households. The bagging method uses Random Forest and Extra Trees, while boosting includes CatBoost and LightGBM. The results showed that the Extra Trees method of bagging provided the best performance with 91% accuracy, 95% F1score, and 97% sensitivity. Factors such as ownership of electronic goods and modern facilities, such as ownership of air conditioners, laptops, and televisions are the most significant variables in influencing the classification of electricity subsidy recipients. With high accuracy and minimal bias, this model effectively supports data-driven policies for electricity subsidy distribution. This research is expected to be a strategic recommendation for the government to improve the effectiveness of the electricity subsidy program to be more efficient, well-targeted, and support the improvement of people's welfare.
PERBANDINGAN ALGORITMA RANDOM FOREST DAN XGBOOST DALAM KLASIFIKASI PENERIMA BANTUAN PANGAN NON-TUNAI (BPNT) DI PROVINSI JAWA BARAT Yulianti, Riska; Ilmani, Erdanisa Aghnia; Waliulu, Megawati Zein; Sartono, Bagus; Firdawanti, Aulia Rizki
Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistika Vol. 6 No. 1 (2025): Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistik
Publisher : LPPM Universitas Bina Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46306/lb.v6i1.850

Abstract

This study compares the performance of Random Forest and XGBoost algorithms in classifying recipients of Non-Cash Food Assistance (BPNT) in West Java Province. The data used is from the 2023 National Socio-Economic Survey (SUSENAS) comprising 25,890 households, with 23.6% BPNT recipients and 76.4% non-recipients. The study includes data exploration, preprocessing, handling class imbalance, baseline modeling, and hyperparameter tuning using Grid Search. The results indicate that undersampling effectively increases the recall of Random Forest to 80.01% and XGBoost to 74.04%, albeit at the expense of accuracy. The most influential variables in classification include the head of household's employment status, flooring material of the house, and type of land/building ownership proof. These findings support the utilization of data-driven algorithms to enhance the accuracy and fairness of BPNT distribution.
Evaluation of Machine Learning Models in Classifying Women's Labor Force Participation in West Java Siregar, Indra Rivaldi; Pratiwi, Windy Ayu; Nugraha, Adhiyatma; Sartono, Bagus; Firdawanti, Aulia Rizki
Techno.Com Vol. 24 No. 1 (2025): Februari 2025
Publisher : LPPM Universitas Dian Nuswantoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62411/tc.v24i1.11945

Abstract

This study compares four classification models—Logistic Regression, Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Adaptive Boosting (AdaBoost)—to predict women's labor force participation in West Java, using a dataset of 62 features. After feature selection, the dataset was reduced to 31 features, followed by modeling with the top 10 most important features from each model. Model performance, evaluated using Balanced Accuracy, F1-Score, and Cohen’s Kappa, showed similar results, with RF and XGBoost slightly outperforming the others. However, the differences were not significant, indicating comparable predictive ability across models. The top 10 features from each model were averaged, and the five most influential features were selected. Key factors influencing women's employment status include household responsibilities, age, education, district minimum wage, and the age of the youngest child. The analysis found that 79.6% of unemployed women manage household duties, while employed women are less involved (18.9%). Age was significant, with employed women mostly in the 35-55 age range, correlating with older children and greater workforce participation. Additionally, employed women are more likely to come from regions with lower minimum wages, suggesting that economic necessity drives their labor market participation. Keywords: female labor force, machine learning, classification, West Java
Optimizing Random Forest Parameters with Hyperparameter Tuning for Classifying School-Age KIP Eligibility in West Java Setyowati, Silfiana Lis; Qalbi, Asyifah; Aristawidya, Rafika; Sartono, Bagus; Firdawanti, Aulia Rizki
Jambura Journal of Mathematics Vol 7, No 1: February 2025
Publisher : Department of Mathematics, Universitas Negeri Gorontalo

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37905/jjom.v7i1.28736

Abstract

Random Forest is an ensemble learning algorithm that combines multiple decision trees to generate a more stable and accurate classification model. This study aims to optimize Random Forest parameters for classifying school-age students' eligibility for the Kartu Indonesia Pintar (KIP) in West Java, based on economic factors. The research uses secondary data from the 2023 National Socio-Economic Survey (SUSENAS) of West Java, with a sample size of 13,044 individuals. To address class imbalance, Synthetic Minority Oversampling Technique (SMOTE) is applied. Hyperparameter tuning through grid search identifies the optimal combination of parameters, including the number of trees (ntree), random variables per split (mtry), and terminal node size (node_size). Model performance is evaluated using balanced accuracy, sensitivity, and specificity. Results indicate that the optimal parameters (mtry = 5, ntree = 674, node_size = 26) yield a balanced accuracy of 65.47%. Significant variables include PKH status, floor area of the house, source of drinking water, and building material type. The model accurately identifies students in need of educational assistance. In conclusion, optimizing Random Forest parameters improves the accuracy of KIP eligibility classification, supporting educational equity policies in West Java. These findings provide a foundation for developing more effective beneficiary selection systems for educational aid.
Evaluasi Kinerja Model Random Forest dan LightGBM untuk Klasifikasi Status Imunisasi Hepatitis B (HB-0) pada Balita Syam, Ummul Auliyah; Irdayanti, Irdayanti; Magfirrah, Indah; Sartono, Bagus; Firdawanti, Aulia Rizki
Euler : Jurnal Ilmiah Matematika, Sains dan Teknologi Volume 13 Issue 1 April 2025
Publisher : Universitas Negeri Gorontalo

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37905/euler.v13i1.29762

Abstract

Hepatitis B (HB-0) immunization in infants is an important step in preventing the transmission of hepatitis B from an early age and improving public health. This study aims to classify the HB-0 immunization status of infants in West Java Province. The method used is the Random Forest and LightGBM algorithms. The research results showed that the Random Forest model had a balanced accuracy of 0.8443, which was slightly higher than LightGBM (0.8357). This indicated that Random Forest performed better in classifying the HB-0 immunization status of infants in West Java Province, accurately distinguishing between those who received and did not receive the immunization without bias toward either class. The global analysis using the Random Forest model identified six feature importance that contributed the most to the model’s performance: BCG immunization status, ownership of the KIA/KMS book, mother’s age, household head’s age, age at first pregnancy, and regency or city classification of residence. The feature importance analysis using SHAP for the first observation showed that BCG immunization status, ownership of the KIA/KMS book, and regency or city classification of residence increased the likelihood of infants receiving immunization. Conversely, the number of children (4), mother’s age (37 years), and household head’s age (40 years) increased the likelihood of infants not receiving immunization. This study is expected to provide data-driven insights for the government to design more effective interventions to improve immunization coverage and child health in Indonesia while also supporting the achievement of global health targets.
Hedging Strategy Analysis of GOTO Stock Using Collar, Bear Put Spread, and Long Strangle Agustiani, Nur; Wahyu, Sri; Firdawanti, Aulia Rizki; Ahmad, Hafidlotul Fatimah
Euler : Jurnal Ilmiah Matematika, Sains dan Teknologi Volume 13 Issue 3 December 2025
Publisher : Universitas Negeri Gorontalo

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37905/euler.v13i3.34092

Abstract

This study compares the performance of three hedging strategies, Collar, Bear Put Spread, and Long Strangle, in a case study of PT GoTo Gojek Tokopedia Tbk (GOTO) stock. The analysis focuses on the risk management effectiveness and profit potential of these strategies within an emerging market context. The research utilizes weekly stock price data from July 2023 to June 2024 (54 observations). The methodological procedures include calculating returns and volatility, testing return normality using the Shapiro-Wilk test, determining European option prices using the Black-Scholes model with a 6% risk-free interest rate, and conducting profit simulations. The findings indicate that the Collar strategy provides maximum protection against stock price declines, albeit with limited profit potential. The Bear Put Spread strategy proves effective in generating returns during moderate price decreases while offering lower risk and cost. Conversely, the Long Strangle strategy possesses high profit potential during significant price volatility but carries the risk of total loss if stock prices remain stagnant. As a comprehensive comparison of these three option strategies applied to GOTO stock, this study recommends the Collar strategy as the optimal choice for risk-averse investors during bearish trends.
Studi Komparatif Metode Boosting Dalam Pengklasifikasian Penerima Bantuan Program Keluarga Harapan (PKH) Amatullah, Fida Fariha; MY, Hadyanti Utami; Rizqi, Tasya Anisah; Wahyuni, Silvia Tri; Sartono, Bagus; Firdawanti, Aulia Rizki
TELKA - Telekomunikasi Elektronika Komputasi dan Kontrol Vol 11, No 3 (2025): TELKA
Publisher : Jurusan Teknik Elektro UIN Sunan Gunung Djati Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15575/telka.v11n3.315-326

Abstract

Ensemble Learning adalah paradigma pembelajaran mesin dimana beberapa model (biasanya disebut "weak learners") dilatih untuk memecahkan masalah yang sama dan digabungkan untuk mendapatkan hasil yang lebih baik. Salah satu model Ensemble, yaitu model boosting. Beberapa metode boosting yang digunakan dalam penelitian ini, yaitu Gradient Boosting Machines (GBM), Extreme Gradient Boosting Machine (XGBM), Light Gradient Boosting Machine (LGBM), dan CatBoost. Penelitian ini akan mengklasifikasikan Rumah Tangga (RT) yang menerima bantuan Program Keluarga Harapan (PKH). Pengklasifikasian PKH sangat penting dilakukan, karena saat ini pemberian PKH belum optimal dan masih banyak yang tidak tepat sasaran. Hasil penelitian menunjukkan bahwa metode LGBM menunjukkan performa terbaik ketika jumlah data latih berukuran besar, yaitu 90% dengan akurasi sebesar 67,97%, sedangkan untuk data latih kecil yaitu 60:40, LGBM memiliki performa yang kurang baik, dengan nilai balanced accuracy terendah dibandingkan metode boosting lainnya, yaitu sebesar 54,43%. Keunggulan LGBM ini disebabkan karena kemampuannya dalam mengelola data besar dan kompleks yang sesuai dengan karakteristik data sosial ekonomi rumah tangga penerima PKH. Dua fitur yang memiliki peran penting untuk pengklasifikasian PKH dalam model terbaik yaitu LGBM adalah faktor ekonomi dan jumlah anggota rumah tangga. Ensemble Learning is a machine learning paradigm in which multiple models (commonly referred to as "weak learners") are trained to solve the same problem and combined to achieve better results. One of the Ensemble models is the boosting model. Several boosting methods used in this study include Gradient Boosting Machines (GBM), Extreme Gradient Boosting Machine (XGBM), Light Gradient Boosting Machine (LGBM), and CatBoost. This study aims to classify households (RT) that receive assistance from the Program Keluarga Harapan (PKH). The classification of PKH recipients is crucial because the distribution of PKH aid has not been optimal, with many cases of misallocation. The results of the study indicate that the LGBM method demonstrates the best performance when the latih dataset is large (90%), achieving an accuracy of 67.97%. However, when the latih dataset is small (60:40), LGBM performs poorly, recording the lowest balanced accuracy among the boosting methods, at 54.43%. The superiority of LGBM is attributed to its ability to handle large and complex data, which aligns with the socio-economic characteristics of PKH recipient households. Two key features that play a significant role in PKH classification using the best-performing model, LGBM, are economic factors and the number of household members.
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.
Co-Authors Ahmad, Hafidlotul Fatimah Alifviansyah, Kevin Amanda, Nabila Tri Amatullah, Fida Fariha Annisa Permata Sari Ardhani, Rizky Aristawidya, Rafika Bagus Sartono Billy Budi Susetyo Cintari, Nanda Putri David Vijanarco Martal Dela Gustiara Denanda Aufadlan Tsaqif Elis Khatizah Fachry Abda El Rahman Gerry Alfa Dito Ghiffary, Ghardapaty Ghaly Hazelita Dwi Rahmasari Hendri Irwandi I Gusti Ngurah Sentana Putra Ilmani, Erdanisa Aghnia Lisa Amelia Magfirrah, Indah Mardatunnisa Isnaini Mega Maulina Mega Ramatika Putri Mirza Farhan Azhari MY, Hadyanti Utami Najib, Mohamad Khoirun Nicholas Abisha Nisa Nur Aisyah Nugraha, Adhiyatma Nur Agustiani Pratiwi, Windy Ayu Qalbi, Asyifah Riska Yulianti, Riska Rizqi Annafi Muhadi Rizqi, Tasya Anisah Rupmana Br Butar Setyowati, Silfiana Lis Shalshabilla Shafa Siregar, Indra Rivaldi Sri Nurdiati Sri Wahyu Syam, Ummul Auliyah Syella Zignora Limba Tsabitah, Dhiya Ulayya Ulfia, Ratu Risha Wahyuni, Silvia Tri Waliulu, Megawati Zein Windi Pangesti Yani Prihantini Hiola Yanuari, Eka Dicky Darmawan Yudhianto, Rachmat Bintang Zulhijrah