Vural Yildirim

Vural Yildirim

Auto-generated profile, based on publicly available information

Vural Yildirim

Institute of Earth and Space Sciences, Eskisehir Technical University, Eskişehir, Turkey

Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 March 2020 in Journal of Statistical Computation and Simulation
Verified
Journal Article
Closed access
VYVural Yildirim
Vural Yildirim
YKYeliz MERT KANTAR
Yeliz MERT KANTAR
Spatial regression models, used to model spatial relationships, have received considerable attention in recent years since numerous data sets have been collected with geographical references in space. The spatial error model (SEM), among spatial regression models, has been widely applied for spatial data in the literature due to its simple structure. However, it is known that the classical estimation methods such as the maximum likelihood and generalized moment can be influenced by the presence of outliers in the data. In this article, a robust estimation approach based on the robustified likelihood equations for SEM is proposed. The results of the simulation study show that the proposed estimator for SEM has smaller bias and mean squared errors and exhibits more robust empirical influence function than the classical methods, when there are outliers in the dataset. The results of all analysis show that the proposed estimator is robust to outliers.

23 May 2014 in Communications in Statistics - Simulation and Computation
Verified
Journal Article
Closed access
YKYeliz Mert Kantar
Yeliz Mert Kantar
VYVural Yildirim
Vural Yildirim
We consider various robust estimators for the extended Burr Type III (EBIII) distribution for complete data with outliers. The considered robust estimators are M-estimators, least absolute deviations, Theil, Siegel's repeated median, least trimmed squares, and least median of squares. Before we perform the aforementioned estimators for the EBIII, we adapt the quantiles method to the estimation of the shape parameter k of the EBIII. The simulation results show that the considered robust estimators generally outperform the existing estimation approaches for data with upper outliers, with certain of them retaining a relatively high degree of efficiency for small sample sizes.

28 March 2023 in Turkish Journal of Remote Sensing and GIS
Verified
Journal Article
Open access
PDF file
VYVural Yildirim
Vural Yildirim
EYErdem Yurdakul
Erdem Yurdakul
GKGökben Adana Karaağaç
Gökben Adana Karaağaç
MKMerve Koçer
Merve Koçer
HUHakan Uyguçgi̇l
Hakan Uyguçgi̇l
Trafik kazalarının önlenmesinde ilk yapılması gereken işlem kazaların yoğunlaştığı noktaların belirlenmesidir. Bu amaçla 2010-2019 yılları arasında Eskişehir kent merkezinde meydana gelen trafik kazaları istatistiksel olarak benzer, yaklaşım olarak farklı iki yöntem kullanılarak analiz edilmiştir. Çalışmada önce klasik sıcak nokta analizi kullanılmış ve 15 sıcak nokta tespit edilmiştir. Daha sonra aynı veri seti konum-zaman küpü kullanılarak zamana bağlı sıcak nokta yöntemi ile analiz edilmiş, 50 aralıklı, 10 yeni, 7 ardışık, 4 sürekli, 1 azalan ve 1 yoğunlaşan olmak üzere toplam 73 sıcak nokta bulunmuştur. İki yöntemin sonuçları kıyaslandığında, zamana bağlı sıcak nokta analizi ile 1. bölgedeki sıcak nokta sayısının 6'dan 19'a, 2. bölgedeki sıcak nokta sayısının 2'den 20'ye, 3. bölgedeki sıcak nokta sayısının 3'den 12'ye, 4. bölgedeki sıcak nokta sayısının 3'den 11'e ve 5. bölgedeki sıcak nokta sayısının 1'den 11'e çıktığı görülmüştür. Klasik sıcak nokta analizine kıyasla zamana bağlı sıcak nokta analizi ile farklı konumlarda ve farklı desenlerde daha çok trafik kazası sıcak noktalarının tespit edilmesi, konumun ve zamanın bir arada kullanılmasının önemini ortaya koymaktadır. Çalışma sonucunda zamana bağlı sıcak nokta analizinin klasik sıcak nokta analizine göre daha detaylı sonuçlar verdiği gözlemlenmiştir.

22 December 2023 in NeuroRehabilitation
Verified
Journal Article
Closed access
PDF file
SÇSamet Çelik
Samet Çelik
VYVural Yıldırım
Vural Yıldırım
ZGZüleyha Damla Güler
Züleyha Damla Güler
HKHüseyin Tuğra Kadam
Hüseyin Tuğra Kadam
BACKGROUND: Computerized neuropsychological tests provide advantages to clinicians with cost, administration, and time. However, studies have pointed out performance differences between manual and computerized versions of some neuropsychological tests. One of these is the Wisconsin Card Sorting Test (WCST). Due to the performance difference, the normative data of manual tests cannot be used for their computerized versions. Therefore, normative data searches are needed for computerized versions. OBJECTIVE: This study aimed to determine the norm values of WCST-CV in a healthy sample. METHODS: 422 healthy adults aged 18–78 participated in this study. WCST-CVsub-scores are modeled by Regression Analysis based on Age and Education level to generate normative data. Among the 13 WCST scores, the regression models for WCST 2, WCST 3, WCST 4, WCST 10, and WCST 11 are significant. WCST 2, WCST 4, and WCST 11 scores are estimated with Ordinary Least Squares (OLS). However, WCST 3 and WCST 10 scores are estimated with Weighted Least Squares (WLS) due to the violation of the homoscedasticity assumption. RESULTS: The regression results show that p-values calculated from error increase as age and education level increase. CONCLUSION: As a result of our research, norm values between 18–78 years of age were produced using RA. It was determined that gender was not significant for any sub-score. Therefore, only age and education level from socio-demographic variables were included in the model.

23 August 2019 in Eskişehir Teknik Üniversitesi Bilim ve Teknoloji Dergisi B - Teorik Bilimler
Verified
Journal Article
Open access
PDF file
VYVural Yildirim
Vural Yildirim
YKYeliz MERT KANTAR
Yeliz MERT KANTAR
The individual pension system is known as the personal future investment system that allows participants to regularly save for their retirements. Thus, the role of this system among incomes in retirement has significantly grown in the last decade in Turkey as in other countries. In Turkey, the participation in the individual pension system has been voluntary since the beginning of this system. Thus, it is important to study the factors affecting the participants of the individual pension system and also the distribution of the participants in Turkey. In this study, participants’ data in the individual pension system of Turkey (2013), provided by the Pension Monitoring Center, is examined at provincial level by means of GIS and spatial statistics tools. Firstly, spatial distribution, spatial local and global clustering for the participation rate have been researched. Also, the participant rate in Turkey with the effects of variables measured at provincial level (the labor rate, unemployment rate, sex ratio, urbanization rate, deposit rate, illiteracy rate and human development index) is studied by using the spatial econometrics models. The result shows that there are global and local spatial autocorrelations between the participation rates in Turkey and also the spatial lag model provides better results than the classical regression model and some other spatial models in terms of all criteria.

6 January 2026 in Wind Engineering
Verified
Journal Article
Closed access
IUİlHan Usta
İlHan Usta
VYVural Yildirim
Vural Yildirim
YKYeliz MERT KANTAR
Yeliz MERT KANTAR
Wind energy is one of the fastest-growing renewable power sources, and its accurate assessment requires statistical distributions capable of capturing the inherent variability in wind speed. This study introduces six power-transformed distributions, Generalized Power Generalized Weibull (GPGWD), Power Generalized Weibull (PGWD), Power Ishita (PID), Power Lindley (PLD), Power Akash (PAD), and Power Shanker (PSD), as new alternatives for wind energy modeling. Using wind speed data from Eskişehir, Türkiye, the proposed models were compared with common benchmark distributions. Results show that GPGWD and PGWD consistently outperform benchmarks, with GPGWD providing the highest accuracy, particularly in R 2 , RMSE, and most KS and CHI tests; PGWD ranks second. Although PSD, PLD, and some other models perform well at stations with irregular or low wind regimes in metrics such as MAPE and PDE, overall findings indicate that GPGWD is the most flexible model across most stations.