@article {10.3844/ajbbsp.2026.22.01.013,
article_type = {journal},
title = {A Prediction Model for Information Anxiety of CSs Based on CMA-ES and XGBoost},
author = {Wang, Bin and Shao, Li},
volume = {22},
number = {1},
year = {2026},
month = {Apr},
pages = {13-1},
doi = {10.3844/ajbbsp.2026.22.01.013},
url = {https://thescipub.com/abstract/ajbbsp.2026.22.01.013},
abstract = {With the rapid advancement of information technology, college students are increasingly facing the problem of information overload. Information anxiety, an emerging mental health issue, significantly impacts their studies, daily lives, and mental well-being. Given the limitations of existing methods in handling complex nonlinear relationships and high-dimensional data, this study proposes an efficient and accurate prediction model for college students’ information anxiety. The model integrates the covariance matrix adaptation evolution strategy (CMA-ES) with the extreme gradient boosting (XGBoost) algorithm to meet the high-precision prediction demands of college mental health early warning systems. CMA-ES, an advanced global optimization algorithm, effectively avoids local optima by dynamically adjusting the covariance matrix, thereby enhancing the model’s global search capabilities and convergence speed. Building on this, the XGBoost algorithm further improves the model’s prediction accuracy and generalization ability through ensemble learning, incorporating hyperparameters optimized by CMA-ES. Experimental results demonstrate that the proposed model outperforms other comparative algorithms in terms of accuracy, recall, F1 score, and area under the curve (AUC). Specifically, the model achieves an accuracy of 0.96, a recall of 0.97, Macro-F1 and Micro-F1 scores of 0.96 and 0.98, respectively, and an AUC close to 1. Additionally, the model maintains a prediction accuracy between 83.1% and 88.5% under varying experimental conditions, with reasonable execution times. The prediction model, which integrates the covariance matrix adaptation evolution strategy and the extreme gradient boosting algorithm, effectively addresses the challenge of accurately predicting information anxiety among college students. It provides a scientific basis and efficient tools for psychological health interventions in universities, holding significant application value and promotion potential.},
journal = {American Journal of Biochemistry and Biotechnology},
publisher = {Science Publications}
}