Research Goal
The goal is to enhance the effective decision-making and intelligent control of complex engineering systems through data analytics, artificial intelligence, and optimization. The methodology includes machine learning, deep learning optimization, transfer learning, dynamic optimization, heuristics, etc.
Head and neck cancer diagnosis from CT scans
Attention-deficit/hyperactivity disorder diagnosis based on fMRI brain data
Breast cancer diagnosis
Cell segmentation and counting
Medical image analysis for 2D and 3D OCT data
Deep learning for thyroid tissue classification and segmentation
Retinal disease diagnosis
Lung nodule diagnosis on CT scan
Prior to joining Mississippi State University
Abnormal event detection in screening printing of surface mount technology
Cleaning cycle prediction
Prior to joining Mississippi State University
Central fill pharmacy system simulation
Association rule-based planogram optimization in robotic medication dispensing system
Robust replenishment planning of pharmacy robotic dispensing systems
Rice diseases detection through deep learning
Machine learning-based 3D micro-flow volume construction
Main Research Directions
- Data analytics for sensitive decision making processes
- Methods: Parallel/Nonparallel support vector machine, ensemble learning, parameter tuning, dimension reduction, noise reduction, deep learning, recurrent neural network, online learning, anomaly event detection, temporal network, spatio-temporal networks
- Applications: Intelligent healthcare, smart manufacturing, surface mount technology, additive manufacturing, social networks mining, human brain functional connectivity pattern analysis (fMRI), power load prediction
- Deep learning and image analysis
- Methods: Convolutional neural networks, generative adversarial networks, neural architecture search, morphological image processing, image synthesis, image denoising, texture analysis, object detection, semantic recognition, image segmentation, 2D/3D image reconstruction
- Applications: Computer-aided diagnosis, autonomous vehicle, machine vision-based defects detection
- Large-scale complex systems optimization
- Methods: Discrete event simulation, agent-based modeling, meta-heuristics, multi-objective evolutionary algorithms, mixed-integer programming, receding horizon control, real-time optimization
- Applications: Warehouse management, supply chain optimization, mail-order pharmacy automation, robotics, planogram optimization, replenishment optimization
Acknowledgement
Mississippi Department of Employment Security
University of Mississippi Medical Center
Integrated Electronics Engineering Center (IEEC) and Watson Institute for Systems Excellence (WISE) at Binghamton University
Department of Systems Science and Industrial Engineering at Binghamton University
