AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Reinforcement Learning with Combinatorial Actions: An Application to Vehicle Routing

Arthur Delarue, Ross Anderson, Christian Tjandraatmadja

NeurIPS 2020 · Reinforcement learning › Policy gradient and actor-critic · approach: Regularized scalable deep value learning · 46 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use RLCA-16 (MIP action selection, 16-hidden-unit value net) instead of greedyCVRP · random CVRP n=51mean total distance: 13.55 -> 10.65Improved
use RLCA-16 instead of Nazari et al.CVRP · random CVRP n=21mean total distance: 6.40 -> 6.16Improved
RLCA-16 versus Kool et al.CVRP · random CVRP n=51mean total distance: 10.62 -> 10.65No clear effect
RLCA-16 versus OR-ToolsCVRP · random CVRP n=51mean total distance: 10.47 -> 10.65Hurt
retain data across policy iterations (retention factor gamma near 1) (ablation)RLCA · CVRP n=11 and n=21gap to optimalImproved
increase sample paths per iteration (ablation)RLCA · CVRP n=11 and n=21gap to optimalImproved
add hidden ReLU nodes (4 vs linear model) (ablation)value function network · CVRP n=11 and n=21gap to optimalImproved
add mild LASSO regularization (ablation)value function network · CVRP n=11 and n=21gap to optimalNo clear effect
include combinatorial lower bounds in objective and training (ablation)RLCA · CVRP n=11 and n=21convergence of gapImproved
increase value network from 0 to 16 hidden nodes (ablation)RLCA action selection MIP (Gurobi) · CVRP n=51MIP runtime (s): 0.097 -> 39.3Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy