Boyang Xu
I am a 5th-year Ph.D. student in Data Science affiliated with the School of Computing and Augmented Intelligence (SCAI) at Arizona State University, advised by Prof. Hao Yan.
My research focuses on the intersection of Reinforcement Learning and Generative AI, specifically investigating how high-dimensional generative models such as diffusion processes and flow matching can enhance decision making under uncertainty. I aim to develop novel algorithmic frameworks that leverage these models to improve sample efficiency and robustness in complex, stochastic environments.
I am actively seeking research engineering internships and collaborative research opportunities. If my work resonates with your team's goals, I'd love to connect. Please feel free to reach out via email.
Distributional/Multi-Agent RL
Bayesian Optimization
Generative Models
Uncertainty Quantification
JAX / PyTorch
Multi-Modal Modeling
news
Jul 2026
My paper, "Adaptive Sampling for Automated Post-Disaster Rapid Damage Assessment via Level-Set Cost-Aware Bayesian Optimization", is accepted to SIAM 2026 Conference on Data Mining (SDM). Looking forward to present our paper in Salt Lake City!
May 2026
I will be serving as a Session Chair for the 2026 INFORMS Annual Meeting! I am organizing the session "Reinforcement Learning for Sequential Decision Making in Complex Systems" under the Quality, Statistics, and Reliability (QSR) cluster. My session will focus on the intersection of RL and generative models for decision-making. See you in Nov, San Francisco, CA!
May 2026
My paper, "Path-Coupled Bellman Flows for Distributional Reinforcement Learning", is accepted to Forty-Third International Conference on Machine Learning (ICML). Looking forward to present our paper in Seoul, South Korea.
Mar 2026
I will be the reviewer for NeurIPS 2026. Look forward to see more interesting papers.
Feb 2026
I will be the reviewer for ICML 2026. Look forward to more contributions to the community.
Oct 2025
I have presented my paper "Adaptive Sampling for Automated Post-Disaster Rapid Damage Assessment via Level-Set Cost-Aware Bayesian Optimization" at Quality, Statistics, and Reliability (QSR) Session in 2025 INFORMS (Atlanta, Georgia).
Aug 2025
My collaborated paper, "Multi-modal Generative Modeling of Event Sequences and Time Series for Solar PV Systems" with Jiayu Huang received the Best Conference Paper Award at IEEE CASE 2025 (Los Angeles, California).
Aug 2025
My paper is accepted to Journal of Quality Technology:
"Partially Observable Markov Decision Process Framework for Operating Condition Optimization
Using Real-Time Degradation Signals."
Jun 2024
My collaborated paper is presented at MSEC 2024:"Support-Structure Decision in 3D Printing With Unsupervised Computer Vision."
Jun 2023
My paper "Data Quality Improvement and Geometric Information Recovery for Manufacturing Thermography with Spatial-Temporal Fast Fourier Transform" is selected as Finalist for the DAIS Division Best Student Paper Competition
at IISE 2023 Annual Conference (New Orleans, Louisiana).
Jun 2023
Our poster, "3D Image Segmentation for Internal Defect Identification in LPBF" was accepted by MSEC 2023 (New Brunswick, NJ).
publications
2026
Adaptive Sampling for Automated Post-Disaster Rapid Damage Assessment via Level-Set Cost-Aware Bayesian Optimization
SIAM 2026 Conference on Data Mining · July 2026
We introduce a cost-aware Bayesian optimization framework with level-set estimation that dynamically guides autonomous data collectors, such as Unmanned Aerial Vehicles, toward the most informative regions to enable rapid, reliable, and cost-effective post-disaster damage assessment.
Path-Coupled Bellman Flows for Distributional Reinforcement Learning
Forty-Third International Conference on Machine Learning (ICML) · May 2026
We introduce path-coupled Bellman flows, a distributional RL method that uses shared-noise coupling between source and target return distributions to reduce variance in flow-matching-based critics.
2025
Multi-Modal Generative Modeling of Event Sequences and Time Series for Solar PV Systems
IEEE International Conference on Automation Science and Engineering (CASE) · August 2025
🏆 Best Conference Paper Award, IEEE CASE 2025
A multi-modal generative framework combining diffusion models with event sequence modeling for solar PV fault diagnosis and prediction.
Partially Observable Markov Decision Process Framework for Operating Condition Optimization Using Real-Time Degradation Signals
Journal of Quality Technology · August 2025
Data Quality Improvement and Geometric Information Recovery for Resistance Spot Welding with Spatial-Temporal Fast Fourier Transform
Quality Engineering · May 2025
🏆 Finalist in DAIS Division Best Student Paper Competition, IISE 2023
2024
Support-Structure Decision in 3D Printing With Unsupervised Computer Vision
International Manufacturing Science and Engineering Conference (MSEC) · August 2024
Three-Dimensional X-Ray Computed Tomography Image Segmentation and Point Cloud Reconstruction for Internal Defect Identification in Laser Powder Bed Fused Parts
Journal of Manufacturing Science and Engineering (JMSE) · June 2024
2023
Defect Segmentation From X-Ray Computed Tomography of Laser Powder Bed Fusion Parts: A Comparative Study Among Machine Learning, Deep Learning, and Statistical Image Thresholding Methods
International Manufacturing Science and Engineering Conference (MSEC) · September 2023