 Depression is a serious mental health condition characterized by persistent sadness or loss of interest in activities, difficulty concentrating, fatigue, changes in appetite, sleep disturbances, and suicidal ideation. Current diagnostic tools for depression are limited and often subjective, making it difficult to accurately identify and treat the disorder. Electroencephalography, e.g., is a promising tool for detecting depression due to its ability to objectively measure brain activity. However, existing e.g. feature extraction methods have not been able to capture the intrinsic patterns of e.g. signals, resulting in poor performance in depression recognition. We proposed a novel regularization parameter-based, improved intrinsic feature extraction method of e.g. signals via Empirical Mode Decomposition, EMD. This method uses an appropriate regularization parameter to generate a regularization matrix, which is then used to calculate the sum of the matrix products of the intrinsic mode functions, IMFs, and the regularization matrix. Finally, the inverse of this matrix is leveraged to extract the intrinsic features. Our method was tested on 4 e.g. datasets and achieved an average accuracy of 0.8750. This article was authored by Jianxian, Yanan Zhang, Hua Jinliang, and others. We are article.tv, links in the description below.