We present a new approach, Programmatic Interpretable reinforcement learning for traffic signal control (π-light), designed to autonomously discover non-differentiable programs.
First-pass extracted concept
LIGHT
Candidate: toolkit item2 source documents7 linked claims
Live refresh every 5sNext refresh in 5s
Aliases
Programmatic Interpretable reinforcement learning for traffic signal control, TNFSF14, π-light
Evidence Snippets
LIGHT (TNFSF14) has emerged as an important modulator of critical innate and adaptive immune responses.
Supporting Sources
Linked Claims
Extensive experiments demonstrate that π-Light consistently outperforms baseline approaches.
Quoted textsource-backed
Extensive experiments demonstrate that our method consistently outperforms baseline approaches.
Learned program policies from π-Light can directly deploy on edge devices with extremely limited resources.
Quoted textsource-backed
Finally, we analyze how the learned program policies can directly deploy on edge devices with extremely limited resources.
π-Light exhibits superior generalization capabilities compared to deep reinforcement learning, enabling training and evaluation across intersections from different cities.
Quoted textsource-backed
Moreover, π-Light exhibits superior generalization capabilities compared to DRL, enabling training and evaluation across intersections from different cities.
π-Light is designed to autonomously discover non-differentiable programs for traffic signal control.
Quoted textsource-backed
We present a new approach, Programmatic Interpretable reinforcement learning for traffic signal control (π-light), designed to autonomously discover non-differentiable programs.
The approach defines a domain specific language and transformation rules for constructing programs, and uses Monte Carlo Tree Search to find an optimal program in a discrete space.
Quoted textsource-backed
Specifically, we define a Domain Specific Language (DSL) and transformation rules for constructing programs, and utilize Monte Carlo Tree Search (MCTS) to find the optimal program in a discrete space.
LIGHT is an important modulator of critical innate and adaptive immune responses.
Quoted textsource-backed
LIGHT (TNFSF14) has emerged as an important modulator of critical innate and adaptive immune responses.
LIGHT, herpesvirus entry mediator, lymphotoxin β receptor, B and T lymphocyte attenuator, and CD160 form an immune regulatory network.
Quoted textsource-backed
LIGHT and its signaling receptors, herpesvirus entry mediator (TNFRSF14), and lymphotoxin β receptor, form an immune regulatory network with two co-receptors of herpesvirus entry mediator, checkpoint inhibitor B and T lymphocyte attenuator, and CD160.