DCIDA uses single-step reinforcement learning with a deterministic action-to-layout mapping to generate distributed resonator circuits that match target transfer functions more closely than prior methods on a surrogate-based benchmark.
A new training approach for parametric modeling of microwave passive components using combined neural networks and transfer functions
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.SY 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Inverse Design in Distributed Circuits Using Single-Step Reinforcement Learning
DCIDA uses single-step reinforcement learning with a deterministic action-to-layout mapping to generate distributed resonator circuits that match target transfer functions more closely than prior methods on a surrogate-based benchmark.