Study Objectives: Obstructive sleep apnea syndrome (OSAS) is a common sleep disorder that occurs in approximately 5-10% of the general population, and characterized by excessive daytime sleepiness, disruptive snoring, recurrent episodes of apnea or hypopnea and nocturnal hypoxemia. The subtypes of positional OSAS (PPs) are defined by conventional classification determined by the apnea-hypopnea index (AHI). However, there were not enough studies about the classification and characterization of PPs in the literature. The aim of this study is to determine the new subtypes of PPs by data mining algorithms. Methods: 'The study was admitted by 514 patients with OSAS with 24 attributes which was analysed by K-means clustering, C&RT and CHAID decision tree algorithms by RStudio programming. Chi-square test was used for cross validation and Kappa statistics were used to compare the re-evaluated values with classical values. Results: In all methods, two clusters for PPs were obtained and the CHAID algorithm gave us the most accurate results. The value for AHI nodes in CHAID was considered as a cut-off value, and cross-validated with the cut-off value obtained by AUC-ROC analysis with high accuracy (92%). Conclusion: It can be concluded that specific treatments should be developed for new subtypes of PPs considering the centroids of 14 significant attributes.