Class Weighting Approach For Handling Imbalanced Data On Forest Fire Classification Using EfficientNet-B1
Wildfires pose significant threats to ecosystems and human safety, necessitating effective monitoring techniques. Detecting forest fires based on images of forest conditions could be a breakthrough. But, the model built from imbalanced data leads to low accuracy. This research addresses the challen...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Universitas Islam Negeri Sunan Kalijaga Yogyakarta
2025-01-01
|
| Series: | JISKA (Jurnal Informatika Sunan Kalijaga) |
| Subjects: | |
| Online Access: | https://ejournal.uin-suka.ac.id/saintek/JISKA/article/view/4831 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832574170758643712 |
|---|---|
| author | Arvinanto Bahtiar Muhammad Ihsan Prawira Hutomo Agung Widiyanto Siti Khomsah |
| author_facet | Arvinanto Bahtiar Muhammad Ihsan Prawira Hutomo Agung Widiyanto Siti Khomsah |
| author_sort | Arvinanto Bahtiar |
| collection | DOAJ |
| description |
Wildfires pose significant threats to ecosystems and human safety, necessitating effective monitoring techniques. Detecting forest fires based on images of forest conditions could be a breakthrough. But, the model built from imbalanced data leads to low accuracy. This research addresses the challenge of class imbalance in multi-class classification for forest fire detection using the EfficientNet-B1 model. This research explores the implementation of class weighting to enhance model performance, particularly focusing on minority classes namely: Fire, Smoke. A dataset of 7,331 training images, categorized into four classes. The results showed that employing the class weighting method achieved an accuracy of 90%. While training duration of 14 minutes and 45 seconds, outperforming the data augmentation method in terms of time efficiency. This study contributes to the development of more effective methods for forest fire monitoring and provides insights for future research in machine learning applications in environmental contexts.
|
| format | Article |
| id | doaj-art-785eb360e4624cd1921b3ea04a4a0e48 |
| institution | Kabale University |
| issn | 2527-5836 2528-0074 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Universitas Islam Negeri Sunan Kalijaga Yogyakarta |
| record_format | Article |
| series | JISKA (Jurnal Informatika Sunan Kalijaga) |
| spelling | doaj-art-785eb360e4624cd1921b3ea04a4a0e482025-02-02T00:37:09ZengUniversitas Islam Negeri Sunan Kalijaga YogyakartaJISKA (Jurnal Informatika Sunan Kalijaga)2527-58362528-00742025-01-01101Class Weighting Approach For Handling Imbalanced Data On Forest Fire Classification Using EfficientNet-B1Arvinanto Bahtiar0Muhammad Ihsan Prawira HutomoAgung WidiyantoSiti KhomsahTelkom University Wildfires pose significant threats to ecosystems and human safety, necessitating effective monitoring techniques. Detecting forest fires based on images of forest conditions could be a breakthrough. But, the model built from imbalanced data leads to low accuracy. This research addresses the challenge of class imbalance in multi-class classification for forest fire detection using the EfficientNet-B1 model. This research explores the implementation of class weighting to enhance model performance, particularly focusing on minority classes namely: Fire, Smoke. A dataset of 7,331 training images, categorized into four classes. The results showed that employing the class weighting method achieved an accuracy of 90%. While training duration of 14 minutes and 45 seconds, outperforming the data augmentation method in terms of time efficiency. This study contributes to the development of more effective methods for forest fire monitoring and provides insights for future research in machine learning applications in environmental contexts. https://ejournal.uin-suka.ac.id/saintek/JISKA/article/view/4831Image Classificationimbalanced dataEfficientNet-B1Forest Fire detection |
| spellingShingle | Arvinanto Bahtiar Muhammad Ihsan Prawira Hutomo Agung Widiyanto Siti Khomsah Class Weighting Approach For Handling Imbalanced Data On Forest Fire Classification Using EfficientNet-B1 JISKA (Jurnal Informatika Sunan Kalijaga) Image Classification imbalanced data EfficientNet-B1 Forest Fire detection |
| title | Class Weighting Approach For Handling Imbalanced Data On Forest Fire Classification Using EfficientNet-B1 |
| title_full | Class Weighting Approach For Handling Imbalanced Data On Forest Fire Classification Using EfficientNet-B1 |
| title_fullStr | Class Weighting Approach For Handling Imbalanced Data On Forest Fire Classification Using EfficientNet-B1 |
| title_full_unstemmed | Class Weighting Approach For Handling Imbalanced Data On Forest Fire Classification Using EfficientNet-B1 |
| title_short | Class Weighting Approach For Handling Imbalanced Data On Forest Fire Classification Using EfficientNet-B1 |
| title_sort | class weighting approach for handling imbalanced data on forest fire classification using efficientnet b1 |
| topic | Image Classification imbalanced data EfficientNet-B1 Forest Fire detection |
| url | https://ejournal.uin-suka.ac.id/saintek/JISKA/article/view/4831 |
| work_keys_str_mv | AT arvinantobahtiar classweightingapproachforhandlingimbalanceddataonforestfireclassificationusingefficientnetb1 AT muhammadihsanprawirahutomo classweightingapproachforhandlingimbalanceddataonforestfireclassificationusingefficientnetb1 AT agungwidiyanto classweightingapproachforhandlingimbalanceddataonforestfireclassificationusingefficientnetb1 AT sitikhomsah classweightingapproachforhandlingimbalanceddataonforestfireclassificationusingefficientnetb1 |