Transformer-based Deep Reinforcement Learning for Psychological Health Evaluation in Traditional Chinese Medicine

• Peng Liu,
• Hongwei Guo,
• Qingming Yu,
• S.P. Raja,

Vol. 19, No. 7, July 31, 2025
10.3837/tiis.2025.07.001, Download Paper (Free):

• Traditional Chinese medicine
• Psychology health
• Evaluation Method
• Transformer
• Deep Reinforcement Learning

Abstract

Investigating the psychological well-being within Traditional Chinese Medicine (TCM) is crucial. It can enhance productivity, improve mental health, reduce occupational hazards, and elevate professional standards. This, in turn, advances the development of TCM. Current TCM psychological health evaluations suffer from inadequate linkage between indicators and outcomes, compromising accuracy. To address this issue, we propose an evaluation method for TCM psychological health with transformer using deep reinforcement learning. First, the overall framework and evaluation indicator system of the evaluation method is designed by combining transformer with deep reinforcement learning algorithms. Second, we establish the mapping between initial indicators and evaluation results through the principles of the transformer, to calculate the comprehensive weight of the evaluation indicator. Finally, the evaluation model is constructed using deep reinforcement learning algorithm, taking the comprehensive weight of the indicator as input and the psychological health status evaluation as output. In addition, we further optimize the model with an adversarial structure of the deep-attention mechanism network, achieving better evaluation results. Experimental results show that our proposed method has good stability, with an evaluation accuracy of over 85% and a misjudgment rate of less than 0.3, and good comprehensive evaluation performance. Therefore, it provides good technical support for the evaluation of psychology health status of TCM.

Statistics

Cite this article