Halil Akinci

Konut pazar değerinin tahmininde makine öğrenmesi algoritmalarının kullanımı, geleneksel istatistiksel modellere kıyasla daha etkili ve başarılı sonuçlar elde etme olanağı sağlamaktadır. Bu çalışmada, XGBoost algoritması kullanılarak Erzurum İli Yakutiye İlçesi örneğinde konut pazar değerinin tahmin edilmesi ve pazar değerini etkileyen başlıca özelliklerin ortaya konulması amaçlanmıştır. Bu doğrultuda, 2024 yılında satışı gerçekleşen ve pazar değerine ulaşılabilen 101 konuttan oluşan veri seti analiz edilmiştir. XGBoost algoritmasıyla geliştirilen model sonuçlarına göre, çalışma alanında konut pazar değeri üzerinde özellikle emlak beyan değeri, net alan (m2), bina yaşı, konut fiyat indeks değeri ve bulunduğu katın belirleyici olduğunu, üniversiteye yakınlık, banyo sayısı, konutun en üst katta bulunması, yönü ve şehir hastanesine yakınlık gibi özelliklerin ise görece daha az etkili olduğu belirlenmiştir. Model performansı, literatürde yaygın olarak kullanılan belirleme katsayısı (R2), ortalama mutlak yüzde hata (MAPE), kök ortalama kare hatası (RMSE) ve ortalama mutlak hata (MAE) metrikleri kullanılarak değerlendirilmiştir. Elde edilen sonuçlara göre modelin R² değeri 0,76, MAPE değeri %17,77, RMSE değeri 515.240,58 ve MAE değeri 405.611,09 olarak hesaplanmıştır. Bu bulgular, çalışma kapsamında ele alınan konut özelliklerinin, konut pazar değerindeki değişimin %76’sını açıkladığını ve XGBoost algoritmasının konut değerleme çalışmalarında kabul edilebilir yeterlilikte performans sağladığını göstermektedir. Bununla birlikte veri setinin genişletilmesiyle başarı performansının daha da artacağı düşünülmektedir. Konutların değer-özellik ilişkisinin mekâna, veri setine ve lokal dinamiklere bağlı olarak değişmesi, çalışmaların farklı coğrafyalarda yapılmasının önemini göstermektedir. Çalışma sonuçlarının yatırımcılar, konut sahipleri, emlakçılar gibi sektörün tüm aktörlerine yön gösterici olması umulmaktadır.

Landslides, which usually occur in mountainous and hilly areas, occur as a result of the soil or rock material forming a slope moving down under the influence of gravity. Forested areas, mostly in mountainous regions, are susceptible to landslides. Forest roads are important infrastructure facilities to protect forest resources and to achieve sustainable management objectives. Forest roads provide many benefits such as facilitating the transportation of wood raw materials, preventing fires and providing access to areas where recreational activities are carried out. However, inappropriately opened forest roads in forest areas cause problems such as landslides, which cause both serious destruction of road networks and serious deformations in forest areas. Landslide-prone forest roads also cause serious economic losses due to disruption of product transport and road maintenance costs. Within the scope of this study, landslide susceptibility maps (LSMs) were produced to determine the relationship between landslides and landslide-causing factors in Handüzü Forest Management Unit of Kastamonu Regional Directorate of Forestry (KRDF) located in the Central Black Sea Region of Türkiye. Land use, altitude, slope, aspect, plan and profile curvature, topographic wetness index (TWI), distance to forest road, drainage networks and fault, crown closure and lithology were used as conditioning factors in the study. Logistic Regression (LR) and Support Vector Machine (SVM) based machine learning models were used to generate LSMs. The receiver operating characteristics (ROC) curve and area under the ROC curve (AUC) method were used to compare the performance of landslide susceptibility models. In the accuracy assessment using the prediction rate curve, the AUC value was 0.968 for the SVM model and 0.668 for the LR model. The AUC values confirmed that SVM performed much better than LR. In addition, the susceptibility of newly planned forest roads (not currently available in the field) in LSMs were determined in the study. As a result of the study, it was determined that the most effective factors affecting landslides in Handüzü Forest Management Directorate are distance to forest roads and drainage networks. In the analyses, it was found that 28.28% of the existing forest roads in the LSM produced with SVM and 56.57% in the LSM produced with LR were found to be in »high« and »very high« landslide susceptible areas. Similarly, 38.43% of the newly planned roads in the LSM produced with SVM and 52.23% in the LSM produced with LR were found to be in »high« and »very high« landslide susceptible areas. These findings showed that forest roads are the main factor in the occurrence of landslides in the study area. Therefore, taking LSMs into account in the planning of forest roads will contribute to reducing the damages that may occur in forest areas due to landslides.

Turkey’s Artvin province is prone to landslides due to its geological structure, rugged topography, and climatic characteristics with intense rainfall. In this study, landslide susceptibility maps (LSMs) of Murgul district in Artvin province were produced. The study employed tree-based ensemble learning algorithms, namely Random Forest (RF), Light Gradient Boosting Machine (LightGBM), Categorical Boosting (CatBoost), and eXtreme Gradient Boosting (XGBoost). LSM was performed using 13 factors, including altitude, aspect, distance to drainage, distance to faults, distance to roads, land cover, lithology, plan curvature, profile curvature, slope, slope length, topographic position index (TPI), and topographic wetness index (TWI). The study utilized a landslide inventory consisting of 54 landslide polygons. Landslide inventory dataset contained 92,446 pixels with a spatial resolution of 10 m. Consistent with the literature, the majority of landslide pixels (70% – 64,712 pixels) were used for model training, and the remaining portion (30% – 27,734 pixels) was used for model validation. Overall accuracy, precision, recall, F1-score, root mean square error (RMSE), and area under the receiver operating characteristic curve (AUC-ROC) were considered as validation metrics. LightGBM and XGBoost were found to have better performance in all validation metrics compared to other algorithms. Additionally, SHapley Additive exPlanations (SHAP) were utilized to explain and interpret the model outputs. As per the LightGBM algorithm, the most influential factors in the occurrence of landslide in the study area were determined to be altitude, lithology, distance to faults, and aspect, whereas TWI, plan and profile curvature were identified as the least influential factors. Finally, it was concluded that the produced LSMs would provide significant contributions to decision makers in reducing the damages caused by landslides in the study area.

Residential real estate is regarded as a safe and profitable investment tool while also meeting the basic human right to housing. The fact that there exists a large number of parameters both affecting the value of a house and varying based on place, person, and time makes the valuation process difficult. In this regard, accurate and realistic price prediction is critical for all stakeholders, particularly purchasers. Machine learning algorithms as an alternative to classical mathematical modeling methods offer great prospects for boosting the efficacy and success rate of price estimating models. Therefore, the purpose of this study is to investigate the applicability and prediction performance of the tree-based ML algorithms -Random Forest (RF), Gradient Boosting Machine (GBM), AdaBoost, and Extreme Gradient Boosting (XGBoost)- in house valuation for Artvin City Center. As a result of the study, the XGBoost and RF algorithms performed the best in estimating house value (0.705 and 0.701, respectively) as determined by the Correlation Coefficients (R2), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE) metrics. Thus, it can be said that ML algorithms, particularly XGBoost and RF, perform satisfactorily in residential real estate appraisal even with modest amounts of data and that the success rate grows as the amount of data increases.

The Eastern Black Sea Region is regarded as the most prone to landslides in Turkey due to its geological, geographical, and climatic characteristics. Landslides in this region inflict both fatalities and significant economic damage. The main objective of this study was to create landslide susceptibility maps (LSMs) using tree-based ensemble learning algorithms for the Ardeşen and Fındıklı districts of Rize Province, which is the second-most-prone province in terms of landslides within the Eastern Black Sea Region, after Trabzon. In the study, Random Forest (RF), Gradient Boosting Machine (GBM), CatBoost, and Extreme Gradient Boosting (XGBoost) were used as tree-based machine learning algorithms. Thus, comparing the prediction performances of these algorithms was established as the second aim of the study. For this purpose, 14 conditioning factors were used to create LMSs. The conditioning factors are: lithology, altitude, land cover, aspect, slope, slope length and steepness factor (LS-factor), plan and profile curvatures, tree cover density, topographic position index, topographic wetness index, distance to drainage, distance to roads, and distance to faults. The total data set, which includes landslide and non-landslide pixels, was split into two parts: training data set (70%) and validation data set (30%). The area under the receiver operating characteristic curve (AUC-ROC) method was used to evaluate the prediction performances of the models. The AUC values showed that the CatBoost (AUC = 0.988) had the highest prediction performance, followed by XGBoost (AUC = 0.987), RF (AUC = 0.985), and GBM (ACU = 0.975) algorithms. Although the AUC values of the models were close to each other, the CatBoost performed slightly better than the other models. These results showed that especially CatBoost and XGBoost models can be used to reduce landslide damages in the study area.

Many large dams built on the Çoruh River have resulted in the inundation of olive groves in Artvin Province, Turkey. This research sets out to identify suitable locations for olive cultivation in Artvin using the random forest (RF) algorithm. A total of 575 plots currently listed in the Farmer Registration System, where olive cultivation is practiced, were used as inventory data in the training and validation of the RF model. In order to determine the areas where olive cultivation can be carried out, a land suitability map was created by taking into account 10 parameters including the average annual temperature, average annual precipitation, slope, aspect, land use capability class, land use capability sub-class, soil depth, other soil properties, solar radiation, and land cover. According to this map, an area of 53,994.57 hectares was detected as suitable for olive production within the study region. To validate the created model, the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC) were utilized. As a result, the AUC value was determined to be 0.978, indicating that the RF method may be successfully used in determining suitable lands for olive cultivation in particular, as well as crop-based land suitability research in general.

Urbanization, which is defined as an irreversible global-scale problem nowadays, necessitates the foundation of new settlement areas. In general, no sufficient scientific assessment and analysis were performed during these processes, and thus, various natural disasters cost the loss of many lives and properties every year. Nevertheless, considering the areas that are risky in terms of natural disasters during the selection of settlement areas might prevent a large-scale loss of lives and properties because of natural disasters. Within the scope of this study, it was aimed to determine suitable settlement areas in the Alanya district, which is one of the significant points of interest for tourists in our country and has a large population and new settlement areas because of this increasing population. Within this scope, besides the risks of flood and landslide that are the most important natural disasters in the region, and a forest fire that is among the most significant risks for the region, also the biocomfort zones were included in the assessments. As a result of the study, it was determined that the most important natural disaster risk was flooding in a large portion of the region and that only 6.72% of the study area was suitable for settlement in terms of all the criteria examined in the present study.