How can models find the right signal?
Feature selection, rough sets, high-dimensional data, efficient learning, and interpretable classification.
Methods · data mining · machine learningPhD Candidate · AI / Data Mining · Industry Researcher
I study how machine intelligence can make complex data more selective, explainable, and useful, then connect those methods to human-centered systems at enterprise scale.
Research agenda
Feature selection, rough sets, high-dimensional data, efficient learning, and interpretable classification.
Methods · data mining · machine learningDelegation, orchestration, explainability, uncertainty, tool permissions, intervention, and decision-centered evaluation.
Agent manager UX · responsible autonomy · evaluationNetwork planning, streaming data, sensing, health applications, and spatial visualization.
Networks · sensing · applied AIPublications + patents
Jin Cao, Chuan Luo, Linlin Xie, Tianrui Li, Hongmei Chen · Boundary-region distance and overlap-degree pre-sorting for efficient feature selection in high-dimensional datasets.
Haowei Jiang, Feiwei Qin, Jin Cao, Yong Peng, Yanli Shao · A computational perspective on recurrent neural network architecture.
Chang Yu, Yongshun Xu, Jin Cao, Ye Zhang, Yinxin Jin, Mengran Zhu · Transformer-based fraud detection evaluated against established machine-learning baselines.
Attention-based classification for imbalanced medical imaging data.
Jin Cao, Yanhui Jiang, Chang Yu, Feiwei Qin, Zekun Jiang · Rough-set methods applied to an interactive therapeutic support system.
Hoang Viet Nguyen, Jin Cao, Guanbo Chen, Boon Loong Ng · US20220137204A1.
Boon Loong Ng, Jianzhong Zhang, Jin Cao, Joonyoung Cho · US11108473B2.
Industry × research
My product work spans enterprise data security, real-time streaming, network intelligence, and agent-assisted workflows. These systems make research questions concrete: tool permissions affect action, latency affects orchestration, uncertainty affects trust, and organizational scale affects whether autonomy can survive contact with reality.
Future teaching interests
These are future teaching interests informed by research and industry practice. I have not held a formal teaching appointment.
Interaction models, evaluation, uncertainty, explainability, guardrails, and accountable automation.
Feature selection, classification, experiment design, visualization, and turning analysis into decisions.
Systems thinking, UX architecture, data governance, AI workflows, and cross-functional specifications.
Academic service
Editorial Board Member
Discover Artificial Intelligence, 2024–present
Editorial Board Member, AI Section
Abdominal Radiology, 2024–present
Research or academic collaboration