Publications
I'm interested in reinforcment learning, robot learning, optimization, and representation learning.
|
|
Pitfall of Optimism: Distributional Reinforcement Learning by Randomizing Risk Criterion
Taehyun Cho, Seungyub Han, Heesoo Lee, Kyungjae Lee, Jungwoo Lee
NeurIPS 2023, 2023
paper /
arxiv /
We provide a perturbed distributional Bellman optimality operator by distorting the risk measure and prove the convergence and optimality of the proposed method with the weaker contraction property.
|
|
SPQR: Controlling Q-ensemble Independence with Spiked Random Model for Reinforcement Learning
Dohyeok Lee, Seungyub Han, Taehyun Cho, Jungwoo Lee
NeurIPS 2023, 2023
paper /
By introducing a novel regularization loss for Q-ensemble independence based on random matrix theory, we propose spiked Wishart Q-ensemble independence regularization (SPQR) for reinforcement learning.
|
|
On the Convergence of Continual Learning with Adaptive Methods
Seungyub Han, Yeongmo Kim, Taehyun Cho, Jungwoo Lee
UAI 2023, 2023
paper /
In this paper, we provide a convergence analysis of memory-based continual learning with stochastic gradient descent and empirical evidence that training current tasks causes the cumulative degradation of previous tasks.
|
|
Learning to Learn Unlearned Feature for Brain Tumor Segmentation
Seungyub Han, Yeongmo Kim, Seokhyeon Ha, Jungwoo Lee, Seunghong Choi
Medical Imaging meets NeurIPS (NeurIPS 2018 Workshop), 2018
paper /
arxiv /
One of the difficulties in medical image segmentation is the lack of datasets with proper annotations. To alleviate this problem, we propose active meta-tune to achieve balanced parameters for both glioma and brain metastasis domains within a few steps.
|
|
Generative Adversarial Trainer: Defense to Adversarial Perturbations with GAN
Hyeungill Lee, Seungyub Han, Jungwoo Lee
, 2017
arxiv /
We propose a novel technique to make neural network robust to adversarial examples using a generative adversarial network. We alternately train both classifier and generator networks. The generator network generates an adversarial perturbation that can easily fool the classifier network by using a gradient of each image.
|
|