Makine Öğrenmesi Yöntemleriyle Orman Yangını Tahmini


Creative Commons License

YILDIRIM O., Günay F. B., Yağanoğlu M.

Iğdır Üniversitesi Fen Bilimleri Enstitüsü Dergisi, cilt.13, sa.3, ss.1468-1481, 2023 (Hakemli Dergi) identifier

Özet

Orman yangını, ormanda yaşama birliği içinde bulunan canlı ve cansız bütün varlıkları yakarak yok eden, ekonomik ve ekolojik zararları olan bir afettir. Son yıllarda küresel ısınma sebebi ile mevsim normalleri üzerinde seyreden sıcaklıklar ve kuraklıklar orman yangını riskini daha da artırmaktadır. Orman yangınları nedeniyle meydana gelen zararı en aza indirmek için yangınla mücadelede erken uyarı, hızlı ve etkin müdahale çok önemlidir. Makine öğrenmesi yöntemleri ise günümüzde erken uyarı sistemlerinde kullanılmaktadır. Bu çalışmada orman yangınlarıyla mücadele için olası orman yangınını önceden tahmin ederek yangınların kontrol edilmesi ve etkisinin azaltılması hedeflenmiştir. Orman yangını tahmin modeli için veri seti, NASA’nın Oak Ridge Ulusal Laboratuvarı (ORNL) Dağıtılmış Aktif Arşiv Merkezi’nin (DAAC) resmi web sitesinden alınarak geliştirilmiştir. Bu veriler makine öğrenmesi yöntemleriyle işlenerek orman yangını tahmin modeli oluşturulmuştur. Veri setine çeşitli ön işleme adımı uygulayarak sınıflandırma modeline uygun hale getirilmiştir. Öznitelik seçme teknikleri ile veri setinin tümü kullanılmadan en yüksek oranda veri bütünlüğü sağlanarak en az sayıda öznitelik alt kümesi seçilmiştir. Hedef değişkeni bulmada en önemli ve en faydalı öznitelikler seçilerek makine öğrenmesi algoritmalarından Destek Vektör Makinesi, Karar Ağacı, Rasgele Orman, Gradyan Artırma, K-En Yakın Komşu ve Naive Bayes olmak üzere 6 farklı sınıflandırma algoritmaları ile model oluşturulmuştur. Model performansını değerlendirmek için validasyon işlemi ve en iyi parametre seçimi için ise hiperparametre optimizasyonu yapılmıştır. Bu çalışmada kullanılan sınıflandırma algoritmaları arasında validasyon işlemi ile birlikte en başarılı iki algoritmadan Rasgele Orman ile %97 ve Naive Bayes ile %96 doğruluk oranı elde edilmiştir.
Forest fire is a disaster that destroys all living and non-living beings in the unity of life in the forest by burning and has economic and ecological damages. In recent years, temperatures and droughts that have been above the seasonal norms due to global warming have increased the risk of forest fires. In order to minimize the damage caused by forest fires, early warning, fast and effective intervention is very important in firefighting. Machine learning methods are used in early warning systems today. In this study, it is aimed to control and reduce the effects of fires by predicting possible forest fires in order to fight forest fires. The dataset for the wildfire prediction model was developed from the official website of NASA's Oak Ridge National Laboratory (ORNL) Center for Distributed Active Archives (DAAC). A forest fire prediction model was created by processing these data with machine learning methods. The data set was adapted to the classification model by applying various preprocessing steps. With the feature selection techniques, the least number of feature subsets were selected by providing the highest level of data integrity without using the entire data set. By choosing the most important and useful features in finding the target variable, a model was created with 6 different classification algorithms, namely Support Vector Machine, Decision Tree, Random Forest, Gradient Boosting, K-Nearest Neighbor and Naive Bayes. Validation process was performed to evaluate model performance and hyperparameter optimization was performed for best parameter selection. Among the classification algorithms used in this study, an accuracy rate of 97% was obtained with Random Forest and 96% with Naive Bayes, which is one of the two most successful algorithms with the validation process.