cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Etis Sunandi
Contact Email
esunandi@unib.ac.id
Phone
6281295949261
Journal Mail Official
jsds_statistika@unib.ac.id
Editorial Address
Jl. WR. Supratman Kelurahan Kandang Limun Kota Bengkulu
Location
Kota bengkulu,
Bengkulu
INDONESIA
Journal of Statistics and Data Science
Published by Universitas Bengkulu
ISSN : -     EISSN : 28289986     DOI : https://doi.org/10.33369/jsds
Core Subject : Economy, Science, Education,
Computer Science & IT Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Mathematics Other
Established in 2022, Journal of Statistics and Data Science (JSDS) publishes scientific papers in the fields of statistics, data science, and its applications. Published papers should be research-based papers on the following topics: experimental design and analysis, survey methods and analysis, operations research, data mining, machine learning, statistical modeling, computational statistics, time series, econometrics, statistical education, and other related topics. All papers are reviewed by peer reviewers consisting of experts and academics across universities and agencies. This journal publishes twice a year, which are March and October.
Arjuna Subject : Matematika - Statistik dan Probabilitas
Articles 2 Documents
 1 
Search results for , issue "vol. 5 no. 2 (2026)" : 2 Documents clear
A Comparative Study of Eigenfaces and Fisherfaces Approaches for Facial Recognition in Pakistan Based on Provincial Demographics Muhammad Ali Baig; Muhammad Yaseen
Journal of Statistics and Data Science Vol. 5 No. 2 (2026)
Publisher : UNIB Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33369/jsds.v5i2.43579

Abstract

The field of statistical pattern recognition is concerned with utilizing statistical techniques to find and categorize patterns in data. This field of study has connections to signal interpreting, computer vision, and machine learning. The initial work on statistical pattern recognition was introduced in the 1960s. Later in the 1990s, it grabbed attention in the fields of Bioinformatics, Geography, Wildlife, and Character recognition by using eigenfaces, fisherfaces, KNN, Fuzzy classification, and Cluster analysis. In this paper, we have used two supervised techniques, Eigenfaces and Fisherface, for face recognition. R (version 4.4.1) software was used for this study. Eight hundred face images of the inhabitants of Pakistan (200 face images from each of the 4 provinces) were used. Using the eigenfaces technique, it was found that 20% of images from Balochistan, 10% of images from Khyber Pakhtunkhwa, and 10% of images from Sindh were not correctly identified. In comparison, 100% of images from Punjab were correctly identified, resulting in an overall 10% error rate. By using the Fisher faces technique, it was found that 3% of images from Balochistan, 2% of images from Khyber Pakhtunkhwa, 0.5% of images from Punjab, and 2% of face images from Sindh were not correctly identified, resulting in overall 1.75% error rate. This study illustrates Fisherfaces' higher efficacy for facial recognition accuracy across multiple geographic representations.
Integration of time series and Artificial Intelligence models to analyze and forecast Vietnam banking stock volatility: Integration of time series and Artificial Intelligence models Quang Phung Duy; Hiep Nguyen Duc; Hai Dang Dinh; Chien Doan Quyet
Journal of Statistics and Data Science Vol. 5 No. 2 (2026)
Publisher : UNIB Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33369/jsds.v5i2.45827

Abstract

This study aims to construct and evaluate a forecasting model for bank stock prices in Vietnam, focusing on the TPB stock of Tien Phong Commercial Joint Stock Bank, through the integration of LSTM neural networks with moving average technical indicators (SMA, EMA, WMA) across time periods of 5, 10, and 20 days. The dataset comprises daily closing price series from August 17, 2017, to July 30, 2025, processed via cleaning, normalization, and training/testing set partitioning steps. Models were trained with 100 and 150 epochs, with performance assessed using MSE, RMSE, MAPE, and R² metrics. Results indicate that the LSTM model combined with SMA20 (100 epochs) achieves optimal performance, with MSE equal 103,685.27, RMSE equal 322, MAPE equal 1.67%, and R² equal 88.23%, excelling due to its noise filtering and long-term trend capturing capabilities. Forecasts for the next three months suggest a positive growth trend, with TPB stock prices potentially ranging from 14,500 to 18,500 VND per share. The study enhances the theoretical framework for hybrid deep learning models in financial time series forecasting, while offering practical recommendations for investors to mitigate systemic risks and optimize strategies in Vietnam's banking sector.

Page 1 of 1 | Total Record : 2

 1 

Filter by Year

2026 2026


Reset
Filter By Issues
All Issue Vol. 5 No. 2 (2026) Vol. 5 No. 1 (2026) Vol. 4 No. 2 (2025) Vol. 4 No. 1 (2025) Vol. 3 No. 2 (2024) Vol. 3 No. 1 (2024) Vol. 2 No. 2 (2023) Vol. 2 No. 1 (2023) Vol. 1 No. 2 (2022) Vol. 1 No. 1 (2022)