Article Per Year (5 Year)
p-Index From 2020 - 2025
0.562
P-Index
This Author published in this journals
All Journal BAREKENG: Jurnal Ilmu Matematika dan Terapan Indonesian Journal of Mathematics and Applications
Fajri, Haidar Ahmad
Unknown Affiliation
Computer Science & IT Control & Systems Engineering Economics, Econometrics & Finance Education Energy Engineering Mathematics Mechanical Engineering Physics Transportation Other

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

Found 3 Documents
Search

 1 
Enhancing Energy Consumption Forecasting with a Multi-Model Deep Learning Approach Fajri, Haidar Ahmad; Oleisan, Kirey
Indonesian Journal of Mathematics and Applications Vol. 3 No. 1 (2025): Indonesian Journal of Mathematics and Applications (IJMA)
Publisher : Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21776/ub.ijma.2025.003.01.5

Abstract

 High energy consumption highlights the need for accurate primary energy forecasts to be critical for policy development, resource optimization and sustainable growth. Indonesia, the fourth largest energy-consuming country in Asia-Pacific, will face challenges in managing energy consumption for economic advancement if it does not conduct proper forecasts with large and limited data. Deep learning models, such as Convolutional Neural Network (CNN), Bidirectional Long Short-Term Memory (Bi-LSTM), and Transformer, excel at extracting insights and modelling temporal dependencies with minimal error, making them ideal for energy forecasting. The hybrid CNN-Bi-LSTM-Transformer model leverages complementary strengths: CNN captures initial patterns, Bi-LSTM manages temporal dependencies, and Transformer enhance global relationships. This model outperforms others model, including Linear Regression, CNN, Bi-LSTM, LSTM, CNN-LSTM, CNN-Bi-LSTM, CNN-Transformer, LSTM-Transformer, Bi-LSTM-Transformer, and hybrid CNN-LSTM-Transformer. It achieves a Mean Squared Error (MSE) of $\num{6.0006e-4}$ on train data, $\num{3.4485e-4}$ on test data and computation time of 8.20 minutes from 25 iterations, with 128 units of CNN layer, 150 units of LSTM layer, and four heads of attention in Transformer. The model also reports a Mean Absolute Error (MAE) of $\num{1.4000e-4}$ for training and $\num{1.5000e-4}$ test data and a Mean Absolute Percentage Error (MAPE) of $1.56$\% for train data and $1.75$\% for test data. This model also effectively tracks energy consumption trends with minimal fluctuations, accurately mirroring the original data and avoiding irregular variations, ensuring reliable future predictions in the long- and short-term.
Enhancing Energy Consumption Forecasting with a Multi-Model Deep Learning Approach Fajri, Haidar Ahmad; Oleisan, Kirey
Indonesian Journal of Mathematics and Applications Vol. 3 No. 1 (2025): Indonesian Journal of Mathematics and Applications
Publisher : Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21776/ub.ijma.2025.003.01.5

Abstract

 High energy consumption highlights the need for accurate primary energy forecasts to be critical for policy development, resource optimization and sustainable growth. Indonesia, the fourth largest energy-consuming country in Asia-Pacific, will face challenges in managing energy consumption for economic advancement if it does not conduct proper forecasts with large and limited data. Deep learning models, such as Convolutional Neural Network (CNN), Bidirectional Long Short-Term Memory (Bi-LSTM), and Transformer, excel at extracting insights and modelling temporal dependencies with minimal error, making them ideal for energy forecasting. The hybrid CNN-Bi-LSTM-Transformer model leverages complementary strengths: CNN captures initial patterns, Bi-LSTM manages temporal dependencies, and Transformer enhance global relationships. This model outperforms others model, including Linear Regression, CNN, Bi-LSTM, LSTM, CNN-LSTM, CNN-Bi-LSTM, CNN-Transformer, LSTM-Transformer, Bi-LSTM-Transformer, and hybrid CNN-LSTM-Transformer. It achieves a Mean Squared Error (MSE) of $\num{6.0006e-4}$ on train data, $\num{3.4485e-4}$ on test data and computation time of 8.20 minutes from 25 iterations, with 128 units of CNN layer, 150 units of LSTM layer, and four heads of attention in Transformer. The model also reports a Mean Absolute Error (MAE) of $\num{1.4000e-4}$ for training and $\num{1.5000e-4}$ test data and a Mean Absolute Percentage Error (MAPE) of $1.56$\% for train data and $1.75$\% for test data. This model also effectively tracks energy consumption trends with minimal fluctuations, accurately mirroring the original data and avoiding irregular variations, ensuring reliable future predictions in the long- and short-term.
IMPROVING SUPPORT VECTOR MACHINE PERFORMANCE WITH BINARY GAUSSIAN IMPROVED WHALE OPTIMIZATION ALGORITHM: A CASE STUDY ON DIABETES DATA Fajri, Haidar Ahmad; Ardiyansa, Safrizal Ardana; Anam, Syaiful; Maharani, Natasha Clarrisa; Julianto, Eric
BAREKENG: Jurnal Ilmu Matematika dan Terapan Vol 19 No 4 (2025): BAREKENG: Journal of Mathematics and Its Application
Publisher : PATTIMURA UNIVERSITY

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30598/barekengvol19iss4pp2531-2542

Abstract

Diabetes mellitus is a chronic condition with high blood sugar that can cause severe organ damage, affecting all ages globally. Early diagnosis is crucial for improving patients' quality of life, and machine learning offers a promising approach. The Support Vector Machine (SVM) is effective for classification, but feature selection is essential to enhance the relevance of features. The Whale Optimization Algorithm (WOA) is an optimal method for global feature selection, but it has a drawback-premature convergence, which can lead to suboptimal results. This issue should be addressed by modifying mutation operations, convergence factors, and population initialization, resulting in Binary Gaussian IWOA (BGIWOA). This research focuses on feature selection using BGIWOA, comparing it with Variance Inflation Factor (VIF) using SVM. The result show that BGIWOA is better than VIF and the best configuration BGIWOA’s parameter is with linear kernel. This configuration produces the best accuracy of 95.00%. BGIWOA-SVM demonstrates better accuracy with stable consistency compared to VIF-SVM. The best SVM model achieves average accuracy of 95.62% for training data and 95.58% for validation data, with an accuracy of 93.85% for the test data. This model also yields an average precision of 94.00%, a recall of 91.00%, and an -score of 92.00%. The model was also better than SVM without optimization, which only achieved a training accuracy of 84.25% and a testing accuracy of 81.30%. This model can assist in diagnosing diabetes with accurate and consistent predictions for new data. The results are specific to the diabetes dataset used in this research, so further testing on other binary datasets is necessary to confirm the model's effectiveness and generalizability across different domains and types of data.
Co-Authors Ardiyansa, Safrizal Ardana Julianto, Eric Maharani, Natasha Clarrisa Oleisan, Kirey Syaiful Anam