TY  - CHAP
U1  - Konferenzveröffentlichung
A1  - Uhlmann, Yannick
A1  - Essich, Michael
A1  - Bramlage, Lennart
A1  - Scheible, Jürgen
A1  - Curio, Cristóbal
T1  - Deep reinforcement learning for analog circuit sizing with an electrical design space and sparse rewards
T2  - MLCAD '22: Proceedings of the 2022 ACM/IEEE Workshop on Machine Learning for CAD, 12-13 September 2022, Snowbird, USA
N2  - There is still a great reliance on human expert knowledge during the analog integrated circuit sizing design phase due to its complexity and scale, with the result that there is a very low level of automation associated with it. Current research shows that reinforcement learning is a promising approach for addressing this issue. Similarly, it has been shown that the convergence of conventional optimization approaches can be improved by transforming the design space from the geometrical domain into the electrical domain. Here, this design space transformation is employed as an alternative action space for deep reinforcement learning agents. The presented approach is based entirely on reinforcement learning, whereby agents are trained in the craft of analog circuit sizing without explicit expert guidance. After training and evaluating agents on circuits of varying complexity, their behavior when confronted with a different technology, is examined, showing the applicability, feasibility as well as transferability of this approach.
KW  - analog circuit sizing
KW  - reinforcement learning
KW  - neural networks
Y1  - 2022
SN  - 978-1-4503-9486-4
SB  - 978-1-4503-9486-4
U6  - https://doi.org/10.1145/3551901.3556474
DO  - https://doi.org/10.1145/3551901.3556474
SP  - 21
EP  - 26
S1  - 6
PB  - Association for Computing Machinery
CY  - New York
ER  -