Analysis of Educational Mental Health and Emotion Based on Deep Learning and Computational Intelligence Optimization

Area of Science:

• Educational Psychology
• Computational Intelligence
• Machine Learning

Background:

• Student mental health is crucial for academic success.
• Early detection of psychological pressure and emotional distress is vital.
• Advancements in AI and deep learning enable sophisticated mental health analysis.

Purpose of the Study:

• To develop and validate a mental health stress detection model for students.
• To incorporate emotional analysis capabilities into the detection model.
• To optimize the model using computational intelligence algorithms.

Main Methods:

• Construction of a dedicated psychological and emotional dataset for educational contexts.
• Implementation of a deep learning-based stress detection model.
• Optimization of the model using the Ant Colony Optimization (ACO) algorithm.
• Validation through comparative experimental analysis and loss function evaluation.

Main Results:

• The constructed dataset is robust, with psychological stress tests indicating generally good student health.
• Specific concerns were identified in obsessive-compulsive disorder and interpersonal sensitivity indicators (average > 0.9).
• The Ant Colony Optimization (ACO) algorithm demonstrated superior stability and execution time compared to other methods.
• Emotion analysis achieved high accuracy, with most differences below 3% and specific emotion differences around 7%.

Conclusions:

• The developed model is feasible and effective for detecting student psychological stress and emotional states.
• The Ant Colony Optimization (ACO) significantly enhances model performance.
• Further optimization is recommended to address remaining limitations in emotion analysis accuracy.