HDI Lab

Our lab explores visual understanding of complex data, intelligent decision-making optimization, and trustworthy AI technologies. We aim to effectively analyze complex data generated across diverse domains and develop interactive systems and intelligent frameworks that can be applied to real-world problem solving.

[Visualization] Capability in Building Visual Analytics Platforms for Complex Spatiotemporal and High-Dimensional Data

We design interactive visual analytics environments that enable intuitive exploration of the spatiotemporal context of complex data generated from diverse domains such as urban data, network logs, and moving objects. By leveraging web-based high-performance rendering engines, we build analytical systems that can visualize large-scale real-time data without latency and help users rapidly uncover meaningful patterns hidden within the data.
[Machine Learning & Optimization] Capability in Intelligent Algorithm-Based Complex Optimization and Reinforcement Learning for Decision Support

We apply advanced metaheuristic and AI optimization techniques, including Genetic Algorithms (GA) and Reinforcement Learning (RL), to solve resource allocation and policy optimization problems involving complex constraints. Through agent-based simulation environments, we explore and learn optimal strategies to develop intelligent decision-support systems that can flexibly respond to the complexity of real-world domains.
[AI Reliability] Capability in Human-Centered Explainable AI (XAI) and Model Reliability Verification

We develop core technologies that visually and transparently explain the reasoning behind black-box AI models, while enabling users to directly interact with data and audit logical flaws or biases in real time. In high-stakes domains such as healthcare, transportation, and security, we propose interactive analytical frameworks that allow experts to evaluate and improve not only model performance but also validity and explainability.

Research Summary