Article: Prediction accuracy enhancement of student performance in nursing education using adaptive reinforcement learning Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.88 pp.51 - 72 Abstract: In nursing simulation training, student performance changes dynamically as they acquire skills. Traditional knowledge tracking can only provide passive assessments and cannot actively adjust strategies, resulting in a disconnect between prediction and intervention. This paper proposes an adaptive prediction framework that integrates dynamic knowledge tracking with quality-function learning. A recurrent neural network tracks the temporal evolution of mastery of nursing knowledge points, using learning gain maximisation as a reward, while an agent dynamically optimises the difficulty and sequence of practice tasks. Using a publicly available nursing skills interaction dataset, this framework improved the area under the curve from 0.790 to 0.873 and the accuracy from 0.742 to 0.801. Comparative analysis demonstrates that integrating reinforcement learning with knowledge tracking effectively enhances predictive performance, providing interpretable insights for personalised interventions in nursing simulation education. Inderscience Publishers - linking academia, business and industry through research

Title: Prediction accuracy enhancement of student performance in nursing education using adaptive reinforcement learning

Authors: Ting Liu; Sha Zhang; Xue Zheng; Xianghui Zhang

Addresses: School of Nursing, Anyang Vocational and Technical College, Anyang, 455000, China ' Anyang Second People's Hospital of Anyang, Anyang, 455000, China ' School of Nursing, Anyang Vocational and Technical College, Anyang, 455000, China ' School of Nursing, Anyang Vocational and Technical College, Anyang, 455000, China

Abstract: In nursing simulation training, student performance changes dynamically as they acquire skills. Traditional knowledge tracking can only provide passive assessments and cannot actively adjust strategies, resulting in a disconnect between prediction and intervention. This paper proposes an adaptive prediction framework that integrates dynamic knowledge tracking with quality-function learning. A recurrent neural network tracks the temporal evolution of mastery of nursing knowledge points, using learning gain maximisation as a reward, while an agent dynamically optimises the difficulty and sequence of practice tasks. Using a publicly available nursing skills interaction dataset, this framework improved the area under the curve from 0.790 to 0.873 and the accuracy from 0.742 to 0.801. Comparative analysis demonstrates that integrating reinforcement learning with knowledge tracking effectively enhances predictive performance, providing interpretable insights for personalised interventions in nursing simulation education.

Keywords: nursing education; adaptive reinforcement learning; student performance prediction.

DOI: 10.1504/IJICT.2026.155570

International Journal of Information and Communication Technology, 2026 Vol.27 No.88, pp.51 - 72

Received: 11 Apr 2026
Accepted: 29 May 2026
Published online: 05 Aug 2026 *

Title: Prediction accuracy enhancement of student performance in nursing education using adaptive reinforcement learning

Keep up-to-date