Akıllı ulaşım sistemleri için ileri istatistiksel yöntemler

The use of machine learning techniques and statistical methods for intelligent transportation systems has gained importance recently. In this thesis, two different methods are proposed for clustering and modeling that can serve the needs of the transportation field. Clustering methods are used to group data points quickly and easily for further analysis of clusters. Also, most clustering methods require complex computations or have an iterative procedure that makes the algorithm time-consuming, especially when the data are relatively large. In this thesis, we propose a new clustering method called spatial adaptive clustering (SAC) based on the idea of adaptive cluster sampling (ACS) design. ACS is a sampling method, which is based on neighborhood search on a grid structure, has an adaptive selection process of units and recursively added units reveal the batched individuals easily and quickly. The SAC algorithm forms clusters based on neighborhood search using grid structures and can detect noise points. The performance of the proposed algorithm is evaluated through comparisons with the results from well-known density-based clustering approaches in the literature using real and artificial data sets. Also, hot spots of accident locations in Birmingham/England and Buca/Turkey are investigated using the SAC algorithm. Additionally, to reduce the number of accidents, hotspot locations are examined in terms of the factors causing the accident. Vehicle headway modeling has also been one of the other important topics for traffic signal optimization and flow modeling. In this thesis, we estimate the parameters of Exponentiated Weibull (EW) distribution using the maximum likelihood method under the assumption that all parameters are unknown. We deal with the performance of ranked set sampling and simple random sampling methods by a simulation study in R-software in terms of mean-squared error. Finally, we illustrate the flexibility and usefulness of EW distribution by analyzing simulated data from a real application study in the transportation field.