EEG-Based Differentiation of Forest and Urban Walking Environments Using Machine Learning

Understanding the impact of different environments on neural activity is essential for environmental neuroscience and cognitive research. This study investigated the effects of forest and urban walking on brain activity using electroencephalography (EEG). EEG data were collected from 30 participants before and after walks in both environments. Spectral analysis was used to extract features from Theta, Alpha, Beta, and Gamma frequency bands. Machine learning classifiers, including Random Forest, Support Vector Machine, and Multi-Layer Perceptron, were employed to differentiate EEG patterns. The Low Beta band yielded the highest average classification accuracy (0.78), indicating its effectiveness in distinguishing between environments. This study is the first to analyze the post-walk effects of forest exposure on brain activity. Our findings highlight the distinct neural impacts of walking in natural versus urban settings, supporting the cognitive benefits of nature exposure and demonstrating the potential of EEG and machine learning for environmental neuroscience applications.