Edy Irwansyah
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Edy Irwansyah has a total of 84 co-authors. Here are the top co-authors by number of shared publications:
- Alexander Agung Santoso Gunawan
Computer Science Department, School of Computer Science, Bina Nusantara University, Jakarta, Indonesia
4 shared publications Rokhana D. Bekti
Institut Sains and Teknologi AKPRIND
3 shared publications- Sri Hartati
Universitas Gadjah Mada
3 shared publications - Bayu Kanigoro
Bina Nusantara University
2 shared publications Eka Miranda
School of Information Systems, Bina Nusantara University, Jakarta, Indonesia
2 shared publications- Erna Fransisca Angela Sihotang
Bina Nusantara University
2 shared publications
Latest Publications
- Open Access
- 1 September 2026
- Published by Kyushu University in Evergreen
- Vol. 13 (3) , 1251-1266
Assessing Nature-Based Solutions for Jakarta Flood Risk Reduction through Integrated Existing and Projected Flood Hazard Mapping
- 1 September 2026
- Published by Kyushu University in Evergreen
- Vol. 13 (3) , 1251-1266
Open AccessAssessing Nature-Based Solutions for Jakarta Flood Risk Reduction through Integrated Existing and Projected Flood Hazard Mapping
- 11 August 2025
- Published by Springer Nature in SN Computer Science
- Vol. 6 (6) , 1-14
Hybrid Indoor Positioning System with Fingerprinting and Multilateration GNN Edge Regression
- 11 August 2025
- Published by Springer Nature in SN Computer Science
- Vol. 6 (6) , 1-14
Hybrid Indoor Positioning System with Fingerprinting and Multilateration GNN Edge Regression
- Open Access
- 24 June 2025
- Published by MDPI AG in Preprints.org
Open Sources Geospatial Intelligence Dashboard for Mangrove Monitoring Using Machine Learning and Remote Sensing Data
Mangroves’ role in global climate mitigation has gained increased attention. From 2000 to 2016, about 62% of global mangrove forests were lost. One significant mangrove area in Indonesia is Sembilang National Park, South Sumatra. Despite its importance, this eco-system has suffered major degradation. Between 2009 and 2017, primary mangrove forest area in the park decreased from 83,447 to 70,263 hectares—a loss of 13,184 hectares. The research focuses on the Sembilang National Park in South Sumatra, Indonesia, over the period of 2019–2023. This study developed a mangrove biomass estimation model using two machine learning algorithms: Random Forest (RF) and Gradient Boosting (GB). RF outperformed GB with an R² score of 0.701 vs. 0.601, highlighting its robustness and better generalization. The biomass (AGB) graph displays total biomass carbon in tonnes, with gradient colouring helping to visualize temporal trends. AGB Trend Overview is imple-mented as a comprehensive metrics dashboard, displaying key information on changes in mangrove biomass. The developed open sources dashboard can provide temporal data analysis of man-grove forest condition, including biomass changes and predictions for the coming year, supporting more targeted data-driven decision-making in mangrove forest conservation and management efforts in Sembilang National Park Indonesia.
- 24 June 2025
- Published by MDPI AG in Preprints.org
Open AccessOpen Sources Geospatial Intelligence Dashboard for Mangrove Monitoring Using Machine Learning and Remote Sensing Data
Mangroves’ role in global climate mitigation has gained increased attention. From 2000 to 2016, about 62% of global mangrove forests were lost. One significant mangrove area in Indonesia is Sembilang National Park, South Sumatra. Despite its importance, this eco-system has suffered major degradation. Between 2009 and 2017, primary mangrove forest area in the park decreased from 83,447 to 70,263 hectares—a loss of 13,184 hectares. The research focuses on the Sembilang National Park in South Sumatra, Indonesia, over the period of 2019–2023. This study developed a mangrove biomass estimation model using two machine learning algorithms: Random Forest (RF) and Gradient Boosting (GB). RF outperformed GB with an R² score of 0.701 vs. 0.601, highlighting its robustness and better generalization. The biomass (AGB) graph displays total biomass carbon in tonnes, with gradient colouring helping to visualize temporal trends. AGB Trend Overview is imple-mented as a comprehensive metrics dashboard, displaying key information on changes in mangrove biomass. The developed open sources dashboard can provide temporal data analysis of man-grove forest condition, including biomass changes and predictions for the coming year, supporting more targeted data-driven decision-making in mangrove forest conservation and management efforts in Sembilang National Park Indonesia.
- Open Access
- 3 April 2025
- Published by Engineering, Technology & Applied Science Research in Engineering, Technology & Applied Science Research
- Vol. 15 (2) , 20580-20587
Deep Learning with Semantic Segmentation Approach for Building Rooftop Mapping in Urban Irregular Housing Complexes
This research investigates the application of the Deep Learning (DL) U-Net architecture for building rooftop segmentation in densely populated urban areas with irregular housing patterns. The research explores the effectiveness of two loss functions - Binary Cross Entropy (BCE) and Dice Loss (DLs) - to optimize the segmentation accuracy. The present study utilized Small-Format Aerial Photography (SFAP) images processed into orthophotos with a final ground sampling distance of 5 cm. The study area, located in Bogor, Indonesia, features both regular and irregular housing patterns, making it an ideal testing ground for the segmentation model. The U-Net model, having been utilized EfficientNetB6 as the encoder and having been trained with augmented data, demonstrated stable performance across metrics, such as accuracy, precision, recall, and F1-score. The results show that the DLs function outperformed BCE, achieving an average Intersection over Union (IoU) score of 96.8% compared to the 87% score for BCE, indicating that DLs is more effective for this application. The study further enhances the segmentation results by converting the raster data into a vector format using the Ramer-Douglas-Peucker (RDP) algorithm, which simplifies and smooths the polygonal shapes of the segmented rooftops. The combination of the U-Net, DLs and RDP algorithm provides high accuracy results and high usability of the segmentation outputs in practical applications, such as urban planning and disaster management scenarios where accurate rooftop delineation is critical.
- 3 April 2025
- Published by Engineering, Technology & Applied Science Research in Engineering, Technology & Applied Science Research
- Vol. 15 (2) , 20580-20587
Open AccessDeep Learning with Semantic Segmentation Approach for Building Rooftop Mapping in Urban Irregular Housing Complexes
This research investigates the application of the Deep Learning (DL) U-Net architecture for building rooftop segmentation in densely populated urban areas with irregular housing patterns. The research explores the effectiveness of two loss functions - Binary Cross Entropy (BCE) and Dice Loss (DLs) - to optimize the segmentation accuracy. The present study utilized Small-Format Aerial Photography (SFAP) images processed into orthophotos with a final ground sampling distance of 5 cm. The study area, located in Bogor, Indonesia, features both regular and irregular housing patterns, making it an ideal testing ground for the segmentation model. The U-Net model, having been utilized EfficientNetB6 as the encoder and having been trained with augmented data, demonstrated stable performance across metrics, such as accuracy, precision, recall, and F1-score. The results show that the DLs function outperformed BCE, achieving an average Intersection over Union (IoU) score of 96.8% compared to the 87% score for BCE, indicating that DLs is more effective for this application. The study further enhances the segmentation results by converting the raster data into a vector format using the Ramer-Douglas-Peucker (RDP) algorithm, which simplifies and smooths the polygonal shapes of the segmented rooftops. The combination of the U-Net, DLs and RDP algorithm provides high accuracy results and high usability of the segmentation outputs in practical applications, such as urban planning and disaster management scenarios where accurate rooftop delineation is critical.
- Open Access
- 17 May 2023
- Published by EDP Sciences in E3S Web of Conferences
- Vol. 388, 01036
Alternative Method in Selecting Location of Subsidized Housing Loan in Medan and Deli Serdang, Indonesia
Indonesia is the fourth most populous country in the world. Therefore, the Indonesian government needs to provide a significant amount of housing for Indonesian citizens. In the place and time people live today, housing prices have risen so fast and so high that the lower classes cannot afford to buy their own houses where the government must provide housing for the citizens. To maximize the government's efficiency in making subsidized housing, the government needs to examine many variables to determine which place is most suitable for housing subsidies. Some examples of values that need to be the object of study are to assess the quality of life or life support value of each location such as the local environment, how strategic the place is in terms of how close it is to public facilities. All information provided is not in numeric values. To apply this model to the analysis, it is necessary to carry out the stages of converting values into the same class for each variable. After assessing and converting all the information into a certain value, it can then be calculated to obtain an objective value to determine the most suitable place for subsidized public housing using two types of modelling systems to get the most accurate results.
- 17 May 2023
- Published by EDP Sciences in E3S Web of Conferences
- Vol. 388, 01036
Open AccessAlternative Method in Selecting Location of Subsidized Housing Loan in Medan and Deli Serdang, Indonesia
Indonesia is the fourth most populous country in the world. Therefore, the Indonesian government needs to provide a significant amount of housing for Indonesian citizens. In the place and time people live today, housing prices have risen so fast and so high that the lower classes cannot afford to buy their own houses where the government must provide housing for the citizens. To maximize the government's efficiency in making subsidized housing, the government needs to examine many variables to determine which place is most suitable for housing subsidies. Some examples of values that need to be the object of study are to assess the quality of life or life support value of each location such as the local environment, how strategic the place is in terms of how close it is to public facilities. All information provided is not in numeric values. To apply this model to the analysis, it is necessary to carry out the stages of converting values into the same class for each variable. After assessing and converting all the information into a certain value, it can then be calculated to obtain an objective value to determine the most suitable place for subsidized public housing using two types of modelling systems to get the most accurate results.
- 4 January 2023
- Published by Springer Nature
- p. 65-74
Analyze Mental Health Disorders from Social Media: A Review
- 4 January 2023
- Published by Springer Nature
- p. 65-74
Analyze Mental Health Disorders from Social Media: A Review
- 4 January 2023
- Published by Springer Nature
- p. 728-739
Deep Learning in Damage Assessment with Remote Sensing Data: A Review
- 4 January 2023
- Published by Springer Nature
- p. 728-739
Deep Learning in Damage Assessment with Remote Sensing Data: A Review
- Open Access
- 30 September 2022
- Published by Universitas Bina Nusantara in Engineering, MAthematics and Computer Science Journal (EMACS)
- Vol. 4 (3) , 103-107
Comparing SVM and Naïve Bayes Classifier for Fake News Detection
Fake news has been evolving into a problem that is getting even more challenging. Technology has been misused to spread false information about many things, such as war, pandemics, and the stock market. Unfortunately, this issue is not a big deal for some people without conscious consumption of that news. Hence, being part takes a role in combating the spread of false information using the advancement of technology. This study proposed two methods of machine learning model, Support Vector Machine (SVM) and Naïve Bayes, to classify fake news. Furthermore, to assert the applicability of models by examining news articles dataset which contain two labels, reliable and unreliable news. The higher accuracy is 0.96 using the SVM model
- 30 September 2022
- Published by Universitas Bina Nusantara in Engineering, MAthematics and Computer Science Journal (EMACS)
- Vol. 4 (3) , 103-107
Open AccessComparing SVM and Naïve Bayes Classifier for Fake News Detection
Fake news has been evolving into a problem that is getting even more challenging. Technology has been misused to spread false information about many things, such as war, pandemics, and the stock market. Unfortunately, this issue is not a big deal for some people without conscious consumption of that news. Hence, being part takes a role in combating the spread of false information using the advancement of technology. This study proposed two methods of machine learning model, Support Vector Machine (SVM) and Naïve Bayes, to classify fake news. Furthermore, to assert the applicability of models by examining news articles dataset which contain two labels, reliable and unreliable news. The higher accuracy is 0.96 using the SVM model