A Unified Framework for Heterogeneity, Contamination, and Missing Data in Multivariate Regression
该研究提出了一种统一框架，通过扩展CG-CWM模型处理回归分析中的异质性、污染数据和缺失值问题，使用ECM算法进行参数估计。
0 citationsRead paper
Selected work
Representative Papers
Privacy Preserving Semantic Communications in Wireless Edge Networks with Vision Language Models
本文提出一种基于视觉-语言模型的隐私保护语义通信框架，通过提取和移除敏感信息，并使用物理层密钥加密文本信息，减少无线边缘网络中的隐私泄露。
0 citationsRead paper
Automating Freshman Course Placement and Registration: A Case Study
This work addresses the inefficiencies of Rowan University’s manual course placement and registration process, which became increasingly unsustainable amid a 57% decade-long increase in incoming student enrollment. To overcome this challenge, the authors developed the institution’s first end-to-end automated system that integrates multi-source data—including the Banner Student Information System and academic advisor-maintained Google Sheets—to intelligently group students by major, automatically assign primary and secondary courses, validate real-time capacity and scheduling constraints, and execute bulk registrations. Leveraging the Banner API, cross-platform data integration, and custom registration scripts, the system successfully enrolled over 3,500 new students, saving more than 350 staff hours annually, substantially reducing error rates, and enhancing operational efficiency and resource allocation—thereby enabling administrative personnel to focus on strategic academic advising.
0 citationsRead paper
Beyond Compliance: A Proposed Framework for Ethical Governance of Student Data in Learning Analytics
This study addresses the prevailing overemphasis on legal compliance in student data governance within learning analytics, which often neglects critical ethical dimensions such as fairness, student autonomy, accountability, and educational purpose. To bridge this gap, the work proposes LEAGUE—a six-pillar ethical governance framework encompassing Legitimacy, Equity, Autonomy, Governance, Utility, and Ethics-by-Design—integrating insights from learning analytics, data ethics, and capability-oriented theories of educational justice. Moving beyond conventional compliance paradigms, this framework pioneers the application of value-sensitive design in the field. Through conceptual review, interdisciplinary theoretical synthesis, and a case analysis of an early warning system, the study demonstrates the framework’s feasibility in enhancing transparency, educational meaningfulness, and ethical justifiability, offering a theoretically grounded yet practically actionable pathway for ethically robust learning analytics.
0 citationsRead paper
CRB-Driven Beamforming and Trajectory Optimization for UAV-assisted ISAC System
This work addresses unmanned aerial vehicle (UAV)-assisted integrated sensing and communication (ISAC) systems by jointly optimizing UAV trajectory and beamforming to minimize the time-averaged Cramér–Rao bound (CRB) on angle-of-arrival estimation, subject to power and mobility constraints, while ensuring reliable downlink communication. The study innovatively adopts the CRB as the optimization objective and incorporates null-space projection-based beamforming to suppress interference. A deep reinforcement learning framework is employed to dynamically co-optimize trajectory and beamforming decisions over discrete time steps. Compared to non-UAV-assisted systems, the proposed approach reduces the time-averaged CRB by more than 10% and significantly outperforms baseline schemes such as fixed-trajectory deployment and maximum ratio transmission.
0 citationsRead paper
Recent publications
Latest Papers
A Unified Framework for Heterogeneity, Contamination, and Missing Data in Multivariate Regression
该研究提出了一种统一框架，通过扩展CG-CWM模型处理回归分析中的异质性、污染数据和缺失值问题，使用ECM算法进行参数估计。
0 citationsRead paper
Privacy Preserving Semantic Communications in Wireless Edge Networks with Vision Language Models
本文提出一种基于视觉-语言模型的隐私保护语义通信框架，通过提取和移除敏感信息，并使用物理层密钥加密文本信息，减少无线边缘网络中的隐私泄露。
0 citationsRead paper
Automating Freshman Course Placement and Registration: A Case Study
This work addresses the inefficiencies of Rowan University’s manual course placement and registration process, which became increasingly unsustainable amid a 57% decade-long increase in incoming student enrollment. To overcome this challenge, the authors developed the institution’s first end-to-end automated system that integrates multi-source data—including the Banner Student Information System and academic advisor-maintained Google Sheets—to intelligently group students by major, automatically assign primary and secondary courses, validate real-time capacity and scheduling constraints, and execute bulk registrations. Leveraging the Banner API, cross-platform data integration, and custom registration scripts, the system successfully enrolled over 3,500 new students, saving more than 350 staff hours annually, substantially reducing error rates, and enhancing operational efficiency and resource allocation—thereby enabling administrative personnel to focus on strategic academic advising.
0 citationsRead paper
Beyond Compliance: A Proposed Framework for Ethical Governance of Student Data in Learning Analytics
This study addresses the prevailing overemphasis on legal compliance in student data governance within learning analytics, which often neglects critical ethical dimensions such as fairness, student autonomy, accountability, and educational purpose. To bridge this gap, the work proposes LEAGUE—a six-pillar ethical governance framework encompassing Legitimacy, Equity, Autonomy, Governance, Utility, and Ethics-by-Design—integrating insights from learning analytics, data ethics, and capability-oriented theories of educational justice. Moving beyond conventional compliance paradigms, this framework pioneers the application of value-sensitive design in the field. Through conceptual review, interdisciplinary theoretical synthesis, and a case analysis of an early warning system, the study demonstrates the framework’s feasibility in enhancing transparency, educational meaningfulness, and ethical justifiability, offering a theoretically grounded yet practically actionable pathway for ethically robust learning analytics.
0 citationsRead paper
CRB-Driven Beamforming and Trajectory Optimization for UAV-assisted ISAC System
This work addresses unmanned aerial vehicle (UAV)-assisted integrated sensing and communication (ISAC) systems by jointly optimizing UAV trajectory and beamforming to minimize the time-averaged Cramér–Rao bound (CRB) on angle-of-arrival estimation, subject to power and mobility constraints, while ensuring reliable downlink communication. The study innovatively adopts the CRB as the optimization objective and incorporates null-space projection-based beamforming to suppress interference. A deep reinforcement learning framework is employed to dynamically co-optimize trajectory and beamforming decisions over discrete time steps. Compared to non-UAV-assisted systems, the proposed approach reduces the time-averaged CRB by more than 10% and significantly outperforms baseline schemes such as fixed-trajectory deployment and maximum ratio transmission.
0 citationsRead paper