Boosting Deep Reinforcement Learning Agents with Generative Data Augmentation

Boosting Deep Reinforcement Learning Agents with Generative Data Augmentation

Data augmentation is a promising technique in improving exploration and convergence speed in deep reinforcement learning methodologies. In this work, we propose a data augmentation framework based on generative models for creating completely novel states and increasing diversity. For this purpose, a...

Mô tả đầy đủ

Chi tiết về thư mục
Những tác giả chính: Tasos Papagiannis, Georgios Alexandridis, Andreas Stafylopatis
Định dạng: Bài viết
Ngôn ngữ:English
Được phát hành: MDPI AG 2023-12-01
Loạt:Applied Sciences
Những chủ đề:
data augmentation
deep reinforcement learning
generative models
Arcade Learning Environment
diffusion models
Truy cập trực tuyến:https://www.mdpi.com/2076-3417/14/1/330

Internet

https://www.mdpi.com/2076-3417/14/1/330

Những quyển sách tương tự