Abstarct: Every business deals with employees who voluntarily resign. In other terms, all businesses experience employee turnover or attrition. However, attrition is a concern when experienced, talented employees leave the organization. Therefore, it is essential to identify the items that caused employee dissatisfaction and know the potential employees who are at risk of resignation. The correct decision can eliminate the unfavorable conditions and prevent talent from leaving the organization. Thanks to technology, and the big storage data in organizations, it is possible to extract valuable knowledge from a large databaxse by using the science of data mining. In this research, data mining was applied to identify talented employees and their characteristics, as well as to identify employees with attrition potential. The purpose of this research is to provide a model to predict which employees are exposed to resign and leave the organization. Then by evaluating their performance if they are part of the organization's talents, it is possible to prevent their departure and provide them with a suitable career path. Dimension reduction and feature selection was applied to improve the results of the data mining process. Firstly, four well-known algorithms: decision tree C4.5, K-nearest neighbor (KNN), Naïve Bayesian Classifiers, and Random Forest are used. At the next step, applying Principal Component Analysis (PCA) algorithm with combination of four other classification algorithms led to increasing the accuracy of algorithms. In fact, the suggested method includes two advantages of reducing the number of features and increasing the classification accuracy. The comparison of the resulting models showed that in a balanced data group, the performance of random forest algorithm was more acceptable among other algorithms and showed less error. Moreover, by combining the mention algorithm with the PCA algorithm, the accuracy, and correctness of the model evaluation would be increased. In contrast, the possibility of error in predicting attrition and predicting talented employees would be decreased.