Occupational health-oriented reinforcement learning models for intelligent adjustment of college students’ employment stress

Abstract The stress of work is a significant issue in the occupational health arena for college students because of the uncertainty of employment, evolving labor market expectations and preparation for the world of work. The existing methods are mainly based on psychological assessment instead of intelligent adjustment of stress according to the occupational health. This study is to design an Adaptive Vortex Optimizer (AVO) based Attention-Based Dueling Deep Q-Network (AD-DQN) model for intelligent adjustment of college students'' employment stress based on personalized decision optimization. The Occupational Employment Stress dataset consists of 15,000 student responses from multiple data sources, including university career guidance, health questionnaires, academic profiles, internship experiences, and students'' self-reported stress responses. Imputation using the K-Nearest Neighbor, Min–Max normalization and outlier refinement are used to enhance data quality. Deep Autoencoder-based feature extraction aims to obtain compact representation by capturing salient occupational, behavioral and educational traits and removing redundant information. The proposed model is designed to use extracted features to optimize the learning parameters and exploration strategies using AVO for efficient convergence. An Attention Mechanism (AM) selects important occupational health and employment related stress signals whereas a Dueling DQN is used to construct an intervention policy for an individual by estimating state value and action advantage. Feedback is offered and the RL constantly adjusts recommendations, thereby furthering career readiness and stress regulation. Experiments conducted in Python show superior performance in terms of 0.982 Accuracy, 0.963 Recall, 0.97 AUC, 0.94 F1-Score and 0.91 Kappa when compared to existing models. From these results, the classification reliability and discriminative ability have been improved, balanced prediction performance and consistent decision making have been achieved. AVO-AD-DQN can support university career-counseling services by enabling personalized employment-stress assessment and adaptive stress-management recommendations.

Authors

Institutions

Publication Details

Journal
Discover Artificial Intelligence
Published
2026-11-17
DOI
https://doi.org/10.1007/s44163-026-02626-w
Primary Topic
Student Stress and Coping
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Occupational health-oriented reinforcement learning models for intelligent adjustment of college students’ employment stress

Runlin Wang
1 links
Discover Artificial Intelligence
Student Stress and Coping
article

Occupational health-oriented reinforcement learning models for intelligent adjustment of college students’ employment stress

Runlin Wang
article en

Abstract

Abstract The stress of work is a significant issue in the occupational health arena for college students because of the uncertainty of employment, evolving labor market expectations and preparation for the world of work. The existing methods are mainly based on psychological assessment instead of intelligent adjustment of stress according to the occupational health. This study is to design an Adaptive Vortex Optimizer (AVO) based Attention-Based Dueling Deep Q-Network (AD-DQN) model for intelligent adjustment of college students'' employment stress based on personalized decision optimization. The Occupational Employment Stress dataset consists of 15,000 student responses from multiple data sources, including university career guidance, health questionnaires, academic profiles, internship experiences, and students'' self-reported stress responses. Imputation using the K-Nearest Neighbor, Min–Max normalization and outlier refinement are used to enhance data quality. Deep Autoencoder-based feature extraction aims to obtain compact representation by capturing salient occupational, behavioral and educational traits and removing redundant information. The proposed model is designed to use extracted features to optimize the learning parameters and exploration strategies using AVO for efficient convergence. An Attention Mechanism (AM) selects important occupational health and employment related stress signals whereas a Dueling DQN is used to construct an intervention policy for an individual by estimating state value and action advantage. Feedback is offered and the RL constantly adjusts recommendations, thereby furthering career readiness and stress regulation. Experiments conducted in Python show superior performance in terms of 0.982 Accuracy, 0.963 Recall, 0.97 AUC, 0.94 F1-Score and 0.91 Kappa when compared to existing models. From these results, the classification reliability and discriminative ability have been improved, balanced prediction performance and consistent decision making have been achieved. AVO-AD-DQN can support university career-counseling services by enabling personalized employment-stress assessment and adaptive stress-management recommendations.

Discover Artificial IntelligenceVol. 6(1)
Hunan University (CN), Hunan College of Information (CN), Changsha Institute of Technology (CN)
Good health and well-being, Decent work and economic growth
Student Stress and Coping
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon