About me

I am a fifth-year Ph.D. candidate in Computer Science at Louisiana State University, advised by Prof. Mingxuan Sun. My research lies at the intersection of machine learning systems, trustworthy AI, and applied AI for time-series and cyber-physical systems.

My primary research focuses on LLM-enabled multimodal time-series forecasting, backdoor attacks on transformer models, and AI-driven risk analysis in real-world domains such as autonomous driving. I am particularly interested in designing practical, data-efficient, and reliable learning systems that bridge foundation models with domain-specific constraints.

I am currently seeking Research Scientist, Applied Scientist, or Machine Learning Engineer positions starting in 2026.

Research Highlights

[DASFAA’25, In-submission-1]Time-Series Forecasting & LLMs: Developed MLTA-LLM framework to align time-series tokens with text embeddings, significantly improving forecasting accuracy and interpretability.

[In-submission-2]Retraining-free backdoor attacks on transformer models, analyzing robustness under realistic software supply-chain threat models.

[In-submission-3]AI Risk Analysis: Investigated AI-driven risk analysis in real-world domains, specifically focusing on safety in autonomous driving systems.

[ICC’24, IoTJ’24]Federated Learning: Proposed a joint device and training scheduling algorithm to achieve cost-efficient federated learning in heterogeneous wireless systems.

[JMSE’24]Marine Geomorphology: Designed a deep learning framework for efficient geomorphic feature identification using sidescan sonar data.

Selected Publications

Enhancing Time Series Forecasting via Multi-Level Text Alignment with LLMs. Taibiao Zhao, Xiaobing Chen, Mingxuan Sun. International Conference on Database Systems for Advanced Applications[(DASFAA’25)][PDF][CODE]

News

Feb 2026[Paper] MLTA-LLM: Multi-Level Text Alignment for Interpretable and Uncertainty-Aware Time Series Forecasting with LLMs get a major revision by TKDD’26

Dec 2025[Service] Invited to serve as a reviewer for TKDD

Dec 2025[Service] Invited to serve as a reviewer for KBS

Oct 2025[Service] Invited to serve as a reviewer for ICLR’26

May 2025[Experience] Joined IHPC as a Research Assistant (MLE), working on AI-driven athlete injury analysis.

Apr 2025[Paper] Enhancing Time Series Forecasting via Multi-Level Text Alignment with LLMs accepted by DASFAA’25