Emotion detection by evaluating Electroencephalogram (EEG) signals is an emerging field of study which provides insights about human emotional states by monitoring brain activity, using music therapy and entertainment. Our study aims to bridge a connection between human brain activities and the recognition of emotion using music. The applications in this study involves mental health assessments, emotionally intelligent agents, adaptive learning, pain assessment, patient monitoring, security and surveillance, as well as personalized music recommendations. Traditional EEG-based emotion detection techniques often struggle with the complex and noisy nature of data. Since the received EEG signals are raw bulk data, in our study, we propose the use of actor critic algorithm which allows accurate and real time emotion detection in the presence of musical stimuli. The actor-critic network architecture is a sophisticated framework designed to predict emotional states from EEG features, leveraging the rich, real-time data that it provides about brain activity. In this setup, the actor network is responsible for generating predictions about an individual’s emotional state based on the EEG signals it processes. It employs these signals to make informed guesses about various emotional conditions, such as happiness, sadness, or stress. On the other hand, the critic network plays a crucial role in evaluating the accuracy of these predictions. It assesses how well the actor’s predictions align with actual emotional states, providing a feedback mechanism that is essential for refining and enhancing the actor’s predictive capabilities.