Yachen Kang
Papers
1
Total Citations
10
H-Index
1
About
Yachen Kang is a researcher advancing the frontiers of deep reinforcement learning, with a primary focus on skill transfer and autonomous decision-making. In their highly regarded 2020 work, "Independent Skill Transfer for Deep Reinforcement Learning," Kang tackles a fundamental challenge: how to efficiently reuse and combine learned primitive skills to master complex, high-level tasks. By leveraging entropy as an intrinsic reward to generate diverse, reusable skill sets, Kang’s research demonstrates that new practical skills can emerge from the composition of simpler ones—a breakthrough that significantly reduces the sample complexity and training time in reinforcement learning. Though early in their career, Kang’s contributions have already garnered over 10 citations, signaling growing recognition among peers. Their work is particularly impactful for robotics and game AI, where agents must adapt to novel environments without starting from scratch. Kang’s focus on independent skill transfer offers a scalable pathway toward more intelligent, generalizable agents, making them a promising voice in the next wave of reinforcement learning innovation.
Research Focus
Key Achievements
Top Papers
- 1Independent Skill Transfer for Deep Reinforcement Learning10 citations · 2020