The number of middle school students experiencing psychological crises in the field of cultural education is increasing, with psychological issues primarily manifesting as problems related to adaptability and development, specifically characterized by symptoms such as depression, anxiety, and confusion. To address this, the paper proposes an improved iForest algorithm combined with a decision cluster classifier based on agglomerative kmeans clustering to develop a psychological issue identification model, and outlines an intervention pathway for addressing student psychological issues. In the results of the graded prediction of students’ mental health status, 3.69% of students were found to have Level 1 psychological issues, 11.32% had Level 2 issues, and 15.74% had Level 3 issues. Therefore, it is essential to actively implement student psychological prevention and intervention pathways to maximize the prevention and reduction of psychological crisis incidents among college students.
