Predicting Theta/Alpha Neurofeedback Success through Psychological and Personality Profiles: A Hybrid Approach Using Multilayer Perceptron and Elastic Net Models

Abstract

The present study aimed to identify and analyze the psychological, cognitive, and neurophysiological factors influencing success in theta/alpha neurofeedback training. The research focused on how personality dimensions (Myers–Briggs Type Indicator), impulsivity (UPPS), intelligence quotient (Raven’s Progressive Matrices), and baseline EEG frequency bands relate to neural self-regulation performance. Methods: A quantitative descriptive–analytical design was employed. Data from six healthy participants who completed eight neurofeedback sessions were collected and analyzed using a multilayer perceptron (MLP) neural network and Elastic Net regression implemented in Python. Results: Findings revealed consistent increases across EEG frequency bands, with baseline neurophysiological measures sufficient for predicting training outcomes. The Elastic Net analysis identified the Judging personality trait, impulsivity, and baseline delta power as the most influential predictors of responsiveness. Furthermore, enhanced negative correlations between theta and alpha bands suggested improved cognitive differentiation during training. Conclusion: Neurofeedback responsiveness is a multifaceted phenomenon influenced by both neurophysiological indices and psychological–cognitive factors. These results underscore the importance of integrating psychological profiling with neural data to optimize individualized neurofeedback interventions.