Human response inhibition, the ability to suppress pre-potent behavior that is inappropriate or that is no longer required, is essential for behavior aimed at everyday life goals. In recent decades, researchers have shown a greater interest in inhibiting the response. Response inhibition is considered an operationalization of certain aspects of impulsivity and compulsivity. Impulsivity is commonly defined as a tendency to act by impulses, acts performed immediately and without voluntary control, while compulsivity is the tendency to repeat a specific behavior and to be unable to inhibit the behavior even when it is no longer appropriate. Due to the importance of inhibiting the response in everyday life, many neuropsychological models have been developed to test inhibitory performance. In this experiment, subjects were asked to respond to a visual stimulus (i.e., a square or circular figure), but withheld this response to irrelevant stimuli (i.e., alarm sound). Therefore, in this study, we developed a new classification model of brain activities under response inhibition using functional connectivity. In this proposed model, we classified the electroencephalography (EEG) signals of successful-stop and failed-stop trials. In this study, we used parametric (LDA, QDA) and non-parametric (PARZENDC, KNNC) classification algorithms to investigate the accuracy of this system. First, the EEG signals acquired during the successful-stop and failed-stop trials, after that, we measured the functional connectivity between seven brain regions, included F3-F4, F4-O1, F4-T8, T7-O1, T7-T8, C3-CZ and C4CZ by phase-locked value (PLV) method. These seven regions of interest covered the entire human brain. These functional connectivity values were used as an input feature for the proposed classification model. Our novel classification system achieved 88% accuracy with QDA. These new findings revealed that this model can be considered to classify the EEG signals of successful-stop and failed-stop trials during response inhibition.