Schizophrenia is a major psychiatric disorder which significantly reduces the quality of life. Early treatment is extremely important in order to mitigate the long-term negative effects. Because of that, reliable diagnosis of schizophrenia is of big interest. In this paper, machine learning based diagnostics of schizophrenia is designed. Classification models are applied to event-related potentials (ERPs) calculated from electroencephalo-gram (EEG) records of patients and healthy subjects performing modification of the visual cued Go-NoGo task. The sample consisted of 200 adult individuals, with an age ranging between 18 and 50 years. In order to apply machine learning models various features are extracted from ERPs. Process of feature extraction is parametrized through a special procedure and parameters of this procedure are selected through a grid-search technique along with model hyperparameters. Feature extraction is followed by Sequential Feature Selection transformation in order to prevent overtitting and reduce computational complexity. Support vector machines and Random Forest models are trained on the resulting feature set. Sensitivity and specificity of the best model are 91% and 91.7% respectively.