Pursuing theoretical and algorithmic advances in data-driven decision making under uncertainty

News

[NeurIPS] Analysis of ODE models for accelerated methods via positive semidefinite kernels

[NeurIPS] Analysis of ODE models for accelerated methods via positive semidefinite kernels

The paper “Convergence analysis of ODE models for accelerated first-order methods via positive semidefinite kernels,” co-authored by Jungbin Kim, has been accepted to Advances in Neural Information ...
[CDC] 3 papers

[CDC] 3 papers

The following three papers from our group have been accepted for presentation at the 62nd IEEE Conference on Decision and Control (CDC). Distributionally robust differential dynamic programming ...
[ICML (Oral)] Unifying Nesterov's AGM for convex and strongly convex objective functions

[ICML (Oral)] Unifying Nesterov’s AGM for convex and strongly convex objective functions

The paper "Unifying Nesterov’s accelerated gradient methods for convex and strongly convex objective functions" has been accepted for Oral presentation at the 40th International Conference on Machine ...
[SICON] Wasserstein linear quadratic control

[SICON] Wasserstein linear quadratic control

The paper "Distributional robustness in minimax linear quadratic control with Wasserstein distance" has been published in the SIAM Journal on Control and Optimization. Distributional robustness in minimax ...
[ICRA] Distributionally robust optimization with unscented transform for learning-based control

[ICRA] Distributionally robust optimization with unscented transform for learning-based control

The paper "Distributionally Robust Optimization with Unscented Transform for Learning-Based Motion Control in Dynamic Environments" has been accepted for presentation at the 2023 IEEE International Conference on ...
[NeurIPS] Improved regret analysis for variance-adaptive linear bandits

[NeurIPS] Improved regret analysis for variance-adaptive linear bandits

The paper "Improved regret analysis for variance-adaptive linear bandits and horizon-free linear mixture MDPs," co-authored by Yeoneung Kim and Kwang-Sung Jun, has been accepted to Advances in ...

Research Areas

Stochastic Control and Reinforcement Learning

Optimization for Machine Learning

Safe Autonomy and Risk Management