Evaluation of occupational accidents in forestry in Europe and Turkey by k-means clustering analysis

Abstract

The incidence rate of occupational accidents is an important indicator of occupational safety performance. The aim of this study was to classify the similarities and differences among 23 European Union (EU) countries along with Turkey in terms of various occupational accident evaluation criteria. This was achieved using the k-means clustering method on data from the forestry and logging sector between 2008 and 2017. The occupational accident assessment criteria used in the study include the nonfatal male accident incidence rate, the nonfatal female accident incidence rate, the total (male + female) nonfatal accident incidence rate, and the total (male + female) fatal accident incidence rate. According to the clustering analysis, three clusters were obtained, and Turkey was included in Cluster 2. By evaluating the final cluster center values and the descriptive statistical values in the clusters, it was found that the occupational accident incidence values of the countries in Cluster 2 were in all four categories below the averages of the other two other clusters as well as the 23 EU countries and Turkey considered together. Cluster 1 was above all of the other clusters as well as all the countries considered together in the total nonfatal occupational accident category, and Cluster 3 in the fatal occupational accident category. Studies similar to this one but on an intercontinental basis would provide a good foundation for improving work health and safety legislation in forestry and logging on a global scale.

Explainable Artificial Intelligence to Support Work Safety in Forestry: Insights from Two Large Datasets, Open Challenges, and Future Work

“…This line of research extends to other countries, with notable studies in Poland [8], Slovakia [9,10], Italy [11], and Brazil [12]. Additionally, cross-national comparative research has been conducted [13]. Notably, a considerable segment of this research is centered on understanding accidents involving chainsaw use or motor-manual tree felling [14][15][16].…”

Modeling Height–Diameter Relationship Using Artificial Neural Networks for Durango Pine (Pinus durangensis Martínez) Species in Mexico

“…Also, other variables such as crown width [18], biomass [19], volume [20], forest fire [21], and annual radial growth with competition indices [22] have been studied with different ML algorithms. Occasionally, a clustering analysis based on unsupervised ML has been included to group similar data points together based on their inherent characteristics or similarities [1,[23][24][25]. An unsupervised clustering analysis can identify patterns or structures in datasets to improve the fitted models in forest modeling.…”