Do Hoang Khoi Nguyen

Do Hoang Khoi Nguyen

AI Researcher, Dreamer, Believer

About me

Welcome to my personal website! I am Do Hoang Khoi Nguyen (Nguyen Do), currently a Ph.D. student in Computer and Information Science and Engineering at the University of Florida, where I work in the Adaptive and Optimization Lab under the supervision of Professor My T. Thai. My research brings together ideas from reinforcement learning, combinatorial optimization, generative models, and Mixture-of-Experts architectures, with a growing interest in applying them to real-world systems like photonic-on-chip networks. I’m especially drawn to problems that sit at the boundary between learning and structured decision-making — where models not only need to perform well but also make reliable, explainable decisions under uncertainty. I enjoy designing systems that are both scalable and theoretically sound, where performance isn’t just about high scores, but also about understanding why and how a model behaves. In practice, this has led me to build frameworks that combine RL with generative modeling for tasks like influence maximization, anomaly detection, or controllable chip design, often using ideas like graph neural networks, latent dynamics, or energy-guided inference. I like models that adapt, reason, and can be trusted in dynamic or noisy environments — especially when applied to domains beyond software, such as networks or hardware. Overall, I care about building AI systems that are not only intelligent, but intelligible, and grounded in both theory and practical relevance.

🎓 Education

  • Ph.D. in Computer & Information Science & Engineering, University of Florida, USA (2024 – Now)
  • B.E. in Electronics & Telecommunication Engineering, PTIT, Vietnam (2016 – 2021) – Top 1%

📄 Recent Publications

  • Q-ShiftDP: A Differentially Private Parameter-Shift Rule for Quantum Machine Learning, International Conference on Artificial Intelligence and Statistics (AISTATS 2026) [Top Tier, A*/A Conference] (Co author)

  • Hexaïssa: Standing on giants’ shoulders—Routing the best chess engines with mixture-of-experts and latent reward learning, Association for the Advancement of Artificial Intelligence (AAAI 2026) [Top Tier, A*/A Conference] (Co-First author)

  • Hephaestus: Mixture Generative Modeling with Energy Guidance for Large-scale QoS Degradation International Conference on Neural Information Processing Systems (NeurIPS 2025) [Top Tier, A*/A Conference] (First author)

  • Swift Hydra: Self-Reinforcing Generative Framework for Anomaly Detection International Conference on Learning Representations (ICLR 2025) [Top Tier, A*/A Conference] (First author)

  • REM: Reinforced Multi-Expert Framework for Influence Maximization Association for the Advancement of Artificial Intelligence (AAAI 2025) [Top Tier, A*/A Conference] (Co-First author)

  • MIM-Reasoner: Learning with Theoretical Guarantees for Multiplex Influence Maximization International Conference on Artificial Intelligence and Statistics (AISTATS 2024) [Top Tier, A*/A Conference] (First author)

  • CHARME: A Chain-based RL Approach for Minor Embedding Problem IEEE/ACM Transactions on Quantum Computing, 2024 [Q1 Journal] (Co-First author)

  • Self-Controlling Photonic-on-Chip Networks Nature Scientific Reports, 2023 [Q1 Journal] (First author)

  • Compact, Efficient 2×2 Three-mode Photonic Switch IEEE Access, 2023 [Q1 Journal] (Co-author)

  • Dual-mode Optical Hybrid Device for SOI Platform IEEE Photonics Journal, 2023 [Q1 Journal] (Co-author)

  • Multi-objective Exploration for PPO IEEE ATiGB Conference, 2022 (First author)

  • See full list at: Google Scholar

💼 Work Experience

  • Optimization Lab, University of Florida, USA (Sep 2022 – Now)
    • RL and probabilistic models for multiplex influence maximization
    • Applied GNNs and theoretical guarantees for quantum minor embedding

  • Connected Brain Corporation, Vietnam (Feb 2023 – Sep 2024)
    • NLP & CV research, multilingual machine translation, GPT-based domain chatbots

  • Naver x PTIT AI Lab, Vietnam (Feb 2020 – Oct 2021)
    • Developed VAEs & GANs for unsupervised object detection and benchmarking

  • MobiFone Telecom Internship (Mar 2019 – May 2019)
    • Built models for fire/liveness detection under poor camera conditions

  • Samsung R&D Center Internship (Aug 2018 – Sep 2018)
    • Built ML modules for games and autonomous driving systems

  • AI Photonic Lab, Vietnam (Jan 2018 – Dec 2018)
    • Researched photonic networks, optimized circuits with ML

👨‍🔬 Past & Current Research Projects

  • Self-Controlling Photonic Network-on-Chip (VINIF Funded)
    Trained RL agents to optimize switching efficiency in on-chip photonics

  • AI Design Optimization for Silicon Photonics
    Used RL and generative models to optimize optical phase-shifting components

  • MO-PPO: Multi-Objective Proximal Policy Optimization
    Balanced energy-efficiency, safety, and task reward in dynamic RL settings

  • Deep Generative Signal Enhancement
    Improved quality of fiber-optic signals via chaos masking + generative models

  • Multiplex Influence Maximization using GNNs
    Developed approximation-guaranteed models for influence spread in multiplex networks

📌 Professional Services

  • Conference Reviewer: NeurIPS (2025), ICLR (2025), AISTATS (2024–2025), AAAI (2025)

  • Program Committee: AAAI 2026: Main Conference Track, AI Alignment Track

🌐 Media Coverage