Proposed robustness analysis and defense methods against adversarial attacks across diverse domains including time series, speech, and fault diagnosis. 

• • Kim et al., Fantastic Robustness Measures, NeurIPS, 2023

  • Kim et al., Bridged adversarial training, Neural Networks, 2023

  • Lee et al., GradDiv: Adversarial Robustness of Randomized Neural Networks via Gradient Diversity Regularization, IEEE TPAMI, 2022

  • and more

Lee et al., GradDiv: Adversarial Robustness of Randomized Neural Networks via Gradient Diversity Regularization, IEEE TPAMI, 2022

Developed privacy-preserving and fair AI learning methods by leveraging cutting-edge technologies such as differential privacy, homomorphic encryption, and generative model-based data synthesis to enhance both data protection and algorithmic fairness.

• • Park et al., Diffusion models for differentially private image classification, CVPR, 2024

  • Choi et al., Fair sampling in diffusion models, AAAI, 2024

  • Byun et al., Parameter-free HE-friendly Logistic Regression, NeurIPS, 2021

  • and more

Park et al., Diffusion models for differentially private image classification, CVPR, 2024

Developed generalizable and stable learning algorithms, such as noise reduction and domain-invariant representation learning, along with numerous machine learning models for industrial applications, including churn prediction, recommendation systems, and document embedding.

• • Park & Lee, Stability Analysis of Denoising Autoencoders, IEEE TKDE, 2021

  • Kim et al., Nonlinear Dynamic Projection for Noise Reduction of Dispersed Manifolds, IEEE TPAMI, 2014

  • Lee et al., Dynamic Characterization of Cluster Structures for Robust and Inductive Support Vector Clustering, IEEE TPAMI, 2006

  • and more

Lee et al., Dynamic Characterization of Cluster Structures for Robust and Inductive Support Vector Clustering, IEEE TPAMI, 2006

Developed global optimization algorithms based on multi-basin dynamical systems and established theoretical foundations for the stability analysis of nonlinear dynamic systems.

• • Lee et al., A dynamical trajectory-based methodology for systematically computing multiple optimal solutions of general nonlinear programming problems, IEEE TAC, 2004

  • Lee et al., Theory of stability regions of a class of non-hyperbolic dynamical systems and its applications to constraint satisfaction problems, IEEE TCAS-I, 2003

  • and more

Lee et al., A dynamical trajectory-based methodology for systematically computing multiple optimal solutions of general nonlinear programming problems, IEEE TAC, 2004

Application of AI technologies to portfolio optimization and asset pricing by integrating traditional financial theories such as the Black-Litterman model with large language models (LLMs) and machine learning techniques.

• • Ko & Lee, Can ChatGPT improve investment decisions?, Finance Research Letters (FRL), 2024

  • Ko et al., A privacy-preserving robo-advisory system with the Black-Litterman portfolio model, JIFMIM, 2023

  • Byun et al., A privacy-preserving mean–variance optimal portfolio, FRL, 2023

  • and more

Empirical studies on blockchain market structures, including analysis of digital currencies (CBDC, Stablecoins, NFTs), detection of MEV (Miner Extractable Value), and investigations into events such as the Terra-LUNA collapse.

• • Park et al., MEV detection via GNNs, Future Generation Computer Systems, 2024

  • Lee et al., Terra-LUNA crash analysis, FRL, 2023

  • Lee et al., Atomic cross-chain settlement model for CBDC, Information Sciences, 2022

  • and more