Machine learning techniques based on ranked set sampling

With the rapid increase in data in recent years and the increasing difficulty of analyzing this data, machine learning is used in many fields of study today. Machine learning is a scientific field of study that aims to develop a variety of algorithms and techniques based on the learning logic of the human brain. Machine learning algorithms study events and try to understand how they occur. As a result of these efforts, they gain the ability to generalize with obtained results. For machine learning algorithms to learn information and to evaluate how well they learn; the data set is split into a training and test set. In the literature, this process is performed randomly at a rate determined by the user. In this thesis study, in addition to the method used in the literature, data set splitting is done by Ranked Set Sampling (RSS), Extreme RSS (ERSS), Median RSS (MRSS), Percentile RSS (PRSS) methods. It is aimed to compare the results obtained. Real life data sets selected from different workspaces are split into training and test sets with the specified methods. Machine learning algorithms were trained with training sets and learning achievements were tested with test sets. As a comparison criteria; root mean square error (RMSE) values were used in the regression algorithms, and the accurate classification rate values were used in the classification algorithms.