Article Per Year (5 Year)
p-Index From 2021 - 2026
0.562
P-Index
This Author published in this journals
All Journal Prosiding Seminar Nasional Official Statistics
Bony Parulian Josaphat
Unknown Affiliation
Humanities Computer Science & IT Economics, Econometrics & Finance Social Sciences

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

 1 
Perbandingan Kinerja Algoritma DE dan PSO dalam Optimasi DTVaR pada Return Harga Propana Mont Belvieu Jernita Bintang Yuliani Manalu; Bony Parulian Josaphat
Seminar Nasional Official Statistics Vol 2026 No 1 (2026): Seminar Nasional Official Statistics 2026
Publisher : Politeknik Statistika STIS

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34123/semnasoffstat.v2026i1.2631

Abstract

Dependent Tail Value-at-Risk (DTVaR) is a measure that quantifies expected loss between two quantiles while accounting for dependence among risk variables. Optimizing DTVaR remains challenging due to the complexity of its objective function. This study compares Differential Evolution (DE) and Particle Swarm Optimization (PSO) for DTVaR optimization using return data of Propane Mont Belvieu price as the target risk variable and Brent crude oil price as the risk variable over 2020–2025. Both algorithms are evaluated using objective values, convergence, and validation tests. The results show that both algorithms converge and produce DTVaR estimates. DE achieves the smallest objective function value. For simulated data, the combination (α= 0,95; δ= 0,8) produces the smallest difference for both algorithms, with PSO yielding the smaller value. For real data, DE achieves the smallest difference at (α= 0,95; δ= 0,8), whereas PSO performs best at (α= 0,95; δ= 0,8). These findings indicate that algorithm performance depends on evaluation criteria and data characteristics.
Estimasi Indeks Pembangunan Manusia Tingkat Kecamatan Menggunakan Data Geospasial Remote Sensing, Point of Interest, dan Statistik Resmi dengan Metode Machine Learning Samuel Marvelius Sianturi; Bony Parulian Josaphat
Seminar Nasional Official Statistics Vol 2026 No 1 (2026): Seminar Nasional Official Statistics 2026
Publisher : Politeknik Statistika STIS

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34123/semnasoffstat.v2026i1.2677

Abstract

The limited data coverage for measuring the Human Development Index (IPM), like National Socio-Economic Survey (Susenas) causes the Statistics Indonesia (BPS) only publishes the IPM at the district/city level. Meanwhile, assessing the quality of human development would be more effective if the IPM were available at a smaller area, such as the sub-district. This study examines the use of geospatial data and official statistics to estimate the IPM at the sub-district level. The results indicate that the best machine learning model is Random Forest Regression with Root Mean Square Error (RMSE) evaluation value of 3.7817, Mean Absolute Error (MAE) of 2.6509, and Mean Absolute Percentage Error (MAPE) of 4.01%. The estimation results show that sub-districts with high estimated HDI are consentrated in areas with high population densities and better accessibility to health, education, and economic facilities, such as sub-districts in Kupang City.
Otomatisasi Narasi Statistik pada Dashboard Publikasi BPS Menggunakan Table-to-Text Generation Juan Lubvrin Fernandez Sidabukke; Bony Parulian Josaphat
Seminar Nasional Official Statistics Vol 2026 No 1 (2026): Seminar Nasional Official Statistics 2026
Publisher : Politeknik Statistika STIS

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34123/semnasoffstat.v2026i1.2679

Abstract

The preparation of descriptive narratives for BPS statistical publications is still performed manually, requiring time and potentially causing errors in numerical interpretation. This study aims to develop PRANATA (Statistical Narrative Automation Portal) to automate statistical narrative generation and integrate it with data visualization in a single dashboard using a Large Language Model (LLM) with Retrieval-Augmented Generation (RAG) and Chain-of-Thought (CoT). The study adopts the CRISP-ML(Q) framework to evaluate Gemini 3 Flash, GPT-OSS (120B), and Llama 3.3 (70B) using 11 statistical indicators from BPS Pematangsiantar Municipality. Gemini 3 Flash achieved the best factuality performance, with Macro Precision of 94.96%, Macro Recall of 96.44%, Macro F1-Score of 95.47%, and Macro Hallucination Rate of 4.88%. It also achieved Fluency and Completeness scores of 5.69 and 5.60 out of 7. PSSUQ with five BPS employees yielded an overall score of 6.00 out of 7. These results indicate that PRANATA supports integrated statistical information dissemination.
Co-Authors Jernita Bintang Yuliani Manalu Juan Lubvrin Fernandez Sidabukke Samuel Marvelius Sianturi