% در این فایل، عنوان پایان‌نامه، مشخصات خود و چکیده پایان‌نامه را به انگلیسی، وارد کنید.
% توجه داشته باشید که جدول حاوی مشخصات پروژه/پایان‌نامه/رساله، به طور خودکار، رسم می‌شود.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\baselineskip=.6cm
\begin{latin}
\latinuniversity{University of Tabriz}
\latinfaculty{Engineering}
\latinsubject{Department of Engineering}
\latinfield{Mathematical Analysis}
\latintitle{ON CLIMATOLOGICAL FEATURES}
\firstlatinsupervisor{Prof.}
%\secondlatinsupervisor{Second Supervisor}
\firstlatinadvisor{Dr. }
%\secondlatinadvisor{Second Advisor}
\latinname{Saeed }
\latinsurname{Sober}
\latinthesisdate{2017}
\latinkeywords{Cluster Analysis, Hydro-climatological stations' Data, Self-Organized Map, Wavelet-Entropy, Utah}
\en-abstract{Classification of hydro-climatologically homogeneous catchments in groups is one of the most significant issues in water resource management. It has its influence on hydrological studies like: regionalization, prediction in ungauged basins (PUB) problem, model parametrization, etc. And cluster analysis as a conventional statistical method has great abilities for this purpose. \\In this study, as the first step, the homogeneous groups of catchments were identified in four ways: by utilizing conventional clustering methods of K-means and Ward; applying Self-Organized Map (SOM), as an unsupervised method, and finally by employing a newly proposed method of WESOM (Wavelet-Entropy based preprocessed data feed to SOM) on the hydro-climatological data of Utah State hydro-stations' of the US as the study case. \\In another word, to investigate the interactions and relationships of hydro-climatological parameters and cluster analysis's results, the time series were analysed in a preprocessed and none- preprocessed manner. The present study, preprocessing method is based on Wavelet transform and the Entropy value of each sub-signal was used as the clustering signature.\\Finally, the homogeneous groups of catchments were defined so as the in-cluster catchments are hydro-climatologically similar to each other and are near to each other in comparison with other clusters' members.\\In this study the mean silhouette value was used as the cluster numbers validation criterion and the results showed that using the newly proposed method for catchment clustering signatures (time-frequency preprocessing of data and utilizing the Entropy value of each sub-series as the clustering signatures instead of using the whole time series) leads to introducing more hydro-climatologically homogeneous clusters in the study area. Thus the proposed method of clustering in this study (the WE SOM) seems to be applicable as a new clustering analysis signature.}
\latinvtitle
\end{latin}