SeaPearl: a constraint programming solver guided by reinforcement learning

DOI10.1007/978-3-030-78230-6_25MaRDI QIDQ2117242zbMATH OpenOpenAlexFDO

Authors Félix Chalumeau, Ilan Coulon, Quentin Cappart, Louis-Martin Rousseau

Publication date 21 March 2022

Full work available at URL https://arxiv.org/abs/2102.09193

zbMATH Keywords

constraint programmingreinforcement learningsolver design

Mathematics Subject Classification ID

Problem solving in the context of artificial intelligence (heuristics, search strategies, etc.) (68T20) Combinatorial optimization (90C27)

Cites work

• A portfolio optimization model with three objectives and discrete variables
• Benchmarking optimization software with performance profiles.
• Constraint-based scheduling: Applying constraint programming to scheduling problems.
• Deep learning
• Failure-directed search for constraint-based scheduling
• Guiding high-performance SAT solvers with unsat-core predictions
• scientific article; zbMATH DE number 956851 (Why is no real title available?)
• Julia: a fresh approach to numerical computing
• JuMP: a modeling language for mathematical optimization
• Machine learning for combinatorial optimization: a methodological tour d'horizon
• MathOptInterface.jl
• MiniCP
• MiniCP: a lightweight solver for constraint programming
• Principles and Practice of Constraint Programming – CP 2004
• Reinforcement learning. An introduction
• Strong combination of ant colony optimization with constraint programming optimization
• The vehicle routing problem
• Three-dimensional matching instances are rich in stable matchings

Cited in

(9)

• SeaPearl
• Learning the travelling salesperson problem requires rethinking generalization
• A learning large neighborhood search for the staff rerostering problem
• Multi-objective optimization for the design of salary structures
• Online learning of variable ordering heuristics for constraint optimisation problems
• Learning and fine-tuning a generic value-selection heuristic inside a constraint programming solver
• Domain-independent dynamic programming
• Enhancing scalability in distributed flexible flowshop scheduling: a hybrid RL-CP approach
• A hybrid learning-based matheuristic to solve the vehicle routing problem with stochastic demands

Describes a project that uses

Uses Software

• MathOptInterface.jl
• MiniCP
• Choco
• OscaR
• NeuroSAT
• Distributions.jl
• VRP
• Julia
• Knet
• Gecode
• JuMP
• LightGraphs.jl

This page was built for publication: SeaPearl: a constraint programming solver guided by reinforcement learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2117242)