İlgi alanları

Sürdürülebilir Kalkınma Amaçları

• 3
• 2
• 2
• 2
• 1
• 1
• 1
• 1

• MakaleYalnız künye2026Scopus SJRQ2/r/bayburt/11185
  Depth-dependent predictability of soil temperature under Mediterranean climatic conditions: a frequency-based perspective

  Guleryuz, Didem; Yıldırım, Ümit; Güler, Cüneyt

  Soil-temperature predictability is inherently depth-dependent because atmospheric forcing is progressively attenuated and transformed into smoother, lower-frequency thermal dynamics with increasing depth. Under Mediterranean climatic conditions, where pronounced seasonality and strong surface forcing coexist, this depth-dependent transformation provides an informative basis for evaluating model behaviour. This study examines soil-temperature prediction at depths of 5, 10, 20, 50, and 100 cm using a multivariate dataset comprising 5,285 retained daily observations from the Mersin meteorological station on Türkiye’s eastern Mediterranean coast. Precipitation, air temperature, relative humidity, wind speed, and evaporation were used as predictors. Support vector regression, random forest, and LightGBM were evaluated using R², RMSE, and MAE. All three models performed strongly in the near-surface layers, with test R² values ranging from 0.948 to 0.954 at depths of 5–20 cm. Predictive performance declined with increasing depth as short-period variability weakened and the thermal signal became smoother and more seasonally structured. RF and LightGBM showed a greater numerical decline at 50–100 cm, whereas SVR achieved test R² values of 0.923 at 50 cm and 0.847 at 100 cm; however, inter-model differences remained modest and were interpreted descriptively. Lomb–Scargle spectral analysis, applied using the actual observation dates, indicated that short-period variability became relatively less pronounced with depth while seasonal-scale and low-frequency components remained prominent. Overall, the findings suggest that soil-temperature prediction should be interpreted in relation to depth-dependent temporal signal characteristics rather than as a search for a universally superior algorithm.

  • 
  İncele
• MakaleAçık Erişim2022Scopus SJRQ2/r/bayburt/13276
  Estimation of soil temperatures with machine learning algorithms—Giresun and Bayburt stations in Turkey

  Guleryuz, Didem

  Since soil temperature (ST) is one of the most critical determinants affecting the soil’s physical and chemical properties, the studies on soil temperature estimation increase with the widespread use of deep learning and machine learning algorithms. This study estimates soil temperature at four depths for Giresun and Bayburt stations in Turkey employing the Bayesian Tuned Gaussian Process Regression (BT-GPR), Bayesian Tuned Support Vector Regression (BT-SVR), and Long Short Term Memory (LSTM) models. The stations were selected from semiarid (Bayburt station) and very humid (Giresun station) climates to compare the models’ performance and measure their applicability in different climate classes. Common meteorological indicators were determined as input parameters in the developed models, and a five-and-a-half-year daily dataset was used for all models. This paper represents a novel scheme to optimize the hyperparameters of kernel functions for GPR and SVR models using the Bayesian optimization method to expand predictive efficiency. The developed GPR and SVR models’ outputs are compared with LSTM via three statistical metrics comprising the root mean square error (RMSE), mean absolute error (MAE), and the coefficient of determination (R2). The results show that the BT-GPR model has a superior estimation ability than other developed models for the two stations. The daily ST estimation with the highest accuracy was obtained at a 5-cm depth using BT-GPR at Giresun station (RMSE = 0.0439, R2 = 0.9535, MAE = 0.0344 in the testing phase) and Bayburt station (RMSE = 0.0525, R2 = 0.9438, MAE = 0.0412 in the testing phase). These outcomes provide helpful benchmarking guidance for future soil temperature investigation at various depths across the selected regions.

  • 
  • 
  • 
  AtıfScopus12·WoS14·OpenAlex21

  İncele
• MakaleYalnız künye2022Scopus SJRQ2/r/bayburt/13241
  Optimized Machine Learning Algorithms for Investigating the Relationship Between Economic Development and Human Capital

  Ozden, Erdemalp; Guleryuz, Didem

  In Economic Development, human capital was previously seen as production factors but gradually evolved into endogenous growth theories. Most of the previous studies have examined the relationships between economic development and human capital via econometric models. Since this relationship is usually nonlinear and machine learning (ML) models can resolve it better, this study investigates the relationships by employing ML methods to provide a new perspective. For this purpose, the optimized ML methods, namely Bayesian Tuned Support Vector Machine and Bayesian Tuned Gaussian Process Regression (BT-GPR), were performed to develop the prediction model for economic development. The hyperparameters have been optimized with the Bayes method by using different kernel functions to increase SVM and GPR methods' predictive performance. The Multiple Linear Regression model has been employed to make a comparison as an econometric model. The performance of the models is evaluated using three statistical metrics, namely, the root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). The BT-GPR with the exponential kernel model has superior prediction ability with the highest accuracy (R2: 0.9727, RMSE: 0.4022, MAE: 0.3728 in the testing phase). The study shows that the BT-GPR model increases the accuracy of R2 6.4%, RMSE 10.7%, and MAE 1% compared with other developed models.

  • 
  AtıfScopus27·WoS18·OpenAlex30

  İncele
• MakaleAçık Erişim2021/r/bayburt/14565
  Türkiye Sağlık Harcamalarının GPR, SVR ve DT Modelleri ile Tahmini

  Guleryuz, Didem

  Ülkeler için artan sağlık maliyetleri ve bu durumun uzun vadeli sürdürülebilirliği siyasi gündeminmerkezinde yer almaktadır. Sağlık harcamalarındaki sürekli artış, hükümet bütçeleri, sağlık hizmetleri vekişisel hasta finansmanı üzerinde baskı oluşturmaktadır. Politika yapıcılar, ortaya çıkabilecek sorunlarauyum sağlamak ve bu maliyetleri düşürmek için reformlar planlamak isterler. Bu durum, planlayıcıları karardestek sistemlerine ve tahmin modellerine yönlendirmiştir. Bu çalışmada, Türkiye’nin Sağlık Harcaması(HS) için çeşitli belirleyicileri dikkate alan bir tahmin modeli tasarlamak amacıyla Destek Vektör Regresyonu(SVR), Regresyon Ağacı (DT) ve Gauss Süreç Regresyonu (GPR) olmak üzere üç makine öğrenmealgoritması kullanılmıştır. Kişi başına gayri safi yurtiçi hasıla, kentsel nüfus oranı, işsizlik oranı, 65 yaş veüstü nüfus, ortalama yaşam süresi, hekim oranı ve toplam hastane yatak sayısı girdi değişkenleri olarakbelirlenmiştir. Veri seti eğitim ve test verisi olarak ayrılmış ve 1990-2019 yılları arası 30 yılı kapsamaktadır.Geliştirilen modeller performans ölçütleri dikkate alınarak karşılaştırılmış ve en iyi model belirlenmiştir.SVR, GPR ve DT modelleri için belirleme katsayısı (R2 ) eğitim aşamasında sırasıyla 0.9929, 0.9989 ve0.9611, test aşamasında sırasıyla 0.9536, 0.8944 ve 0.1166’dır. Ayrıca, SVR modeli, test aşamasında enyüksek R2ve en düşük kök ortalama kare hatası değerleri ile en iyi tahmin sonuçlarına sahiptir. Çalışma,önerilen SVR modelinin RMSE değerini diğer GPR ve DT modellerine kıyasla sırasıyla % 32.02 ve % 39.66azalttığını göstermiştir. Sonuç olarak, Türkiye’nin sağlık harcamaları SVR modeli kullanılarak yüksekdoğrulukta tahmin edilebilir.

  • 
  AtıfOpenAlex10·TR Dizin4

  İncele
• MakaleAçık Erişim2021/r/bayburt/14335
  Clustering framework to cope with COVID-19 for cities in Turkey

  Guleryuz, Didem; Ozden, Erdemalp

  Introduction: This article is the product of the research Clustering Framework to Cope with COVID-19 for Cities in Turkey, developed at Bayburt University in 2021. Problem: Turkey's risk map, presented in January 2021, to take local decisions in tackling the COVID-19 pandemic, was based on confirmed cases only. Health, socio-economic and environmental indicators are also important for management decisions of COVID-19. The risk map to be designed by adding these indicators will support more effective decisions. Objective: The research aims to propose a clustering scheme to design a risk map of cities for Turkey. Methodology: The unsupervised clustering algorithm suggested dividing the cities of Turkey into clusters, considering health, socio-economic, environmental indicators, and the spread pattern of COVID-19. Results: We found that cities are clustered into five groups while megacity Istanbul alone formed a cluster, three of Turkey's largest cities formed another cluster. Other clusters consist of 19, 26, and 32 cities, respectively. The most important determinants which have predictive power are identified. Conclusion: The suggested clustering method can be a decision support system for policymakers to determine the differences and similarities of cities in quarantine decisions and normalization phases for the following periods of the pandemic. Originality: To the best of our knowledge, this study differs from previous studies because countries were grouped in previous studies by only considering the confirmed cases. In this study, cities were clustered in terms of the health, socio-economic, and environmental indicators to make decisions locally. Limitations: The distribution of confirmed cases by age could be added, especially to make decisions about education, but this data is not officially announced.

  • 
  • 
  • 
  AtıfWoS1·OpenAlex1

  İncele
• MakaleAçık Erişim2021Scopus SJRQ1/r/bayburt/13297
  Forecasting outbreak of COVID-19 in Turkey; Comparison of Box–Jenkins, Brown's exponential smoothing and long short-term memory models

  Guleryuz, Didem

  The new coronavirus disease (COVID-19), which first appeared in China in December 2019, has pervaded throughout the world. Because the epidemic started later in Turkey than other European countries, it has the least number of deaths according to the current data. Outbreak management in COVID-19 is of great importance for public safety and public health. For this reason, prediction models can decide the precautionary warning to control the spread of the disease. Therefore, this study aims to develop a forecasting model, considering statistical data for Turkey. Box-Jenkins Methods (ARIMA), Brown's Exponential Smoothing model and RNN-LSTM are employed. ARIMA was selected with the lowest AIC values (12.0342, -2.51411, 12.0253, 3.67729, -4.24405, and 3.66077) as the best fit for the number of total case, the growth rate of total cases, the number of new cases, the number of total death, the growth rate of total deaths and the number of new deaths, respectively. The forecast values of the number of each indicator are stable over time. In the near future, it will not show an increasing trend in the number of cases for Turkey. In addition, the pandemic will become a steady state and an increase in mortality rates will not be expected between 17–31 May. ARIMA models can be used in fresh outbreak situations to ensure health and safety. It is vital to make quick and accurate decisions on the precautions for epidemic preparedness and management, so corrective and preventive actions can be updated considering obtained values.

  • 
  AtıfScopus39·WoS28·OpenAlex47

  İncele
• Kitap BölümüYalnız künye2021/r/bayburt/13210
  Covid-19 Impact on Global Supply Chains

  Güleryüz, Didem

  İncele
• MakaleAçık Erişim2020/r/bayburt/14837
  LSTM ve Facebook Prophet Kullanarak Brent Ham Petrol Trendinin Tahmini

  Guleryuz, Didem; Ozden, Erdemalp

  Ham petrol ve petrol ürünleri, endüstriyel üretimin önemli girdileri arasında olduğu kadar lojistik ve taşımacılıkta da kritik bir roloynamaktadır. Dolayısıyla, petrol fiyatlarındaki ani artışlar ve düşüşler küresel ekonomilerde ve dahası ekonomiler üzerinde doğrudanveya dolaylı bir etkisi vardır. Ayrıca, gelişmekte olan ekonomilerdeki krizler, büyük ekonomiler arasındaki ticaret anlaşmazlıkları vepetrol fiyatının dinamik doğası, petrol arz ve talebi üzerinde etkisi olmaktadır ve petrol fiyatında zaman zaman oynaklık çok sertolmaktadır. Petrol fiyatlarındaki bu belirsizlikler hem tüketicilere hem de üreticilere ağır potansiyel kayıplar yaratabilmektedir. Bu hızlıdeğişkenlik ve dalgalanma nedeniyle petrol fiyatlarının tahmin edilmesi küresel öneme sahiptir. Bu çalışmada, Brent Petrol fiyatlarınıngelecekteki trendini tahmin edilebilmek için geçmiş değerleri girdi alan Uzun Kısa Süreli Bellek (LSTM) ve Facebook Prophet (FBPr)yöntemleri kullanılmıştır. İki modelin petrol fiyatları için Haziran 1988 ile Haziran 2020 arasında haftalık 32 yıllık veri seti kullanılarakkarşılaştırılmış ve en uygun model belirlenmiştir. Veri seti eğitim ve test setleri olmak üzere iki gruba ayrılmıştır; eğitim seti için ilkyirmi beş yıl seçilirken ve son yedi yıl ise tahmin doğruluğunu onaylamak için kullanılmıştır. LSTM ve FBPr modelleri için katsayıtayini (R2 ) eğitim aşamasında 0.92, 0.89 ve test aşamasında 0.89, 0.62 bulunmuştur. Elde edilen sonuçlar incelendiğinde, LSTMmodelinin petrol fiyatlarındaki trendi tahmin etmek için daha iyi sonuç verdiği görülmüştür.

  AtıfOpenAlex14·TR Dizin6

  İncele
• MakaleAçık Erişim2020/r/bayburt/12195
  Evaluation of waste management using clustering algorithm in megacity Istanbul

  Guleryuz, Didem

  Industrialization and urbanization are increasing with the effect of globalization worldwide. The waste management problems are rising with the rising population rate, industrialization, and economic developments in the cities, which turned into environmental problems that directly affect human health. This study aims to examine waste management performance in the districts located in the city of Istanbul. To ensure that the districts are clustered in terms of the similarities and differences base on waste management. On this occasion, the authorized unit managers of the districts in the same cluster will be able to establish similar management policies and make joint decisions regarding waste management. In addition, the division of districts into clusters according to the determining indicators can provide information about the locations of waste storage centers. Also, these clusters will form the basis for the optimization constraints required to design appropriate logistics networks. Waste management performance of 39 districts in Istanbul in 2019 was compared by taking into consideration domestic waste, medical waste, population, municipal budget, and mechanical sweeping area. The data were obtained from The Istanbul Metropolitan Municipality (IMM) and Turkey Statistical Institute (TURKSTAT). One of the non-hierarchical clustering methods, the K-means clustering method, was applied using IBM SPSS Modeler data mining software to determine the relations between 39 districts. As a result, the waste management performance of the districts was evaluated according to the statistical data, similarities and differences were revealed by using the determined indicators.

  • 
  • 
  • 
  AtıfScopus12·OpenAlex11·TR Dizin1

  İncele