Quran memorizing leads to locating humans in a relaxation state, resulting in a slower reaction time (RT). This changes the event time and amplitude of the waves of event-related potentials (ERP). Nevertheless, studies on Quran memorizing have only employed the EEG signals to investigate the effects of Quran memorizing on the life quality of autistic patients and people with stroke. In this study, we accordingly analyzed ERP waves, especially P200, N200, and P300 waves, in two groups of Quran memorizers and non-memorizers recorded during the retrieval phases of pattern recognition memory (PRM) test. In fact, we first selected optimal channels according to the energy topography of ERP signals in the retrieval phase. Then, we extracted a set of features from ERP signals using discrete wavelet transform. Finally, we also evaluated the separability of these ERP signals using multilayer perceptron neural network (MLPNN) and support vector machine developed by the optimal combinations of features selected by T-test analysis, the sequential feature selection (SFS), and the genetic algorithm. The outcome of this analysis generally indicated that MLPNN developed by the optimal features obtained from SFS algorithm had higher accuracy to separate two groups of Quran memorizers and non-memorizers (99.3% and 93.2% for training and testing sets, respectively). Therefore, this finding generally represented that the Quran memorizing as a factor involved in the visual memory can be used as a protocol to change the attention and concentration in humans.