Scopus
🔓 Açık Erişim YÖKSİS Eşleşti
Classification of Forest Fires in European Countries by Clustering Analysis Techniques
Sakarya University Journal of Science · Ekim 2023
YÖKSİS Kayıtları
Classification of Forest Fires in European Countries by Clustering Analysis Techniques
Sakarya University Journal of Science · 2023 TR DİZİN
DOKTOR ÖĞRETİM ÜYESİ MUSLU KAZIM KÖREZ →
Makale Bilgileri
DergiSakarya University Journal of Science
Yayın TarihiEkim 2023
Cilt / Sayfa27 · 987-1001
Scopus ID2-s2.0-85218011495
Erişim🔓 Açık Erişim
Özet
The biggest threat to the forests, which are natural habitats in European countries, as they are in the whole world, is forest fires. The aim of this study is to group the 38 European countries which have completely accessible fire indexes between the years 2008 to 2022; with respect to their similarities in fire regimes; and to compare the obtained groups with respect to their fire indexes. The clustering technique, which is a data mining method, was used while making these comparisons since it would be more objective and realistic to group and evaluate the countries according to their similarities. In the K-Means technique 2 clusters, and in the Ward's method 3 clusters were obtained. In the K-Means technique, significant statistical differences were found between the 2 clusters in terms of all fire indexes (p<0.05). In the Ward’s method, statistically significant differences were found between the clusters in terms of the number of fires, total area burned (ha) and woodland (p<0.05). In the result of the studies, the fire regimes in Turkey, Bosnia and Herzegovina, Ukraine, Italy, Spain, and Portugal resulted higher than the other countries in both clustering algorithms. Since many factors were taken into consideration in the study, countries heavily associated with fires such as Greece and France were separated from those with high fire regimes. It is recommended to conduct modelling studies with data mining algorithms by taking different fire indexes into account in order to increase the reliability of the results.
Yazarlar (4)
1
Hakan Serin
2
M. K. Korez
ORCID: 0000-0001-9524-6115
3
Mehmet Emin Tekin
4
Sinan Siren
ORCID: 0000-0003-2182-5047
Anahtar Kelimeler
cluster analysis
Data mining method
forest fire
k-means
ward method
Kurumlar
Selçuk Tip Fakültesi
Konya Turkey
Selçuk Üniversitesi
Selçuklu Turkey