Combinatorial Optimization and Reasoning with Graph Neural Networks

Cappart, Quentin; Chételat, Didier; Khalil, Elias B.; Lodi, Andrea; Morris, Christopher J.; Veličković, Petar

doi:10.24963/ijcai.2021/595

Cited by 83 publications

(37 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results indicate that GNNs have become increasingly appealing alternatives in solving combinatorial problems. Moreover, there are also some well-organized literature reviews on this subject, such as (Bengio, Lodi, and Prouvost 2021), (Cappart et al 2021) and.…”

Section: Related Workmentioning

confidence: 99%

Can Graph Neural Networks Learn to Solve MaxSAT Problem?

Liu¹,

Jia²,

Huang³

et al. 2021

Preprint

View full text Add to dashboard Cite

With the rapid development of deep learning techniques, various recent work has tried to apply graph neural networks (GNNs) to solve NP-hard problems such as Boolean Satisfiability (SAT), which shows the potential in bridging the gap between machine learning and symbolic reasoning. However, the quality of solutions predicted by GNNs has not been well investigated in the literature. In this paper, we study the capability of GNNs in learning to solve Maximum Satisfiability (MaxSAT) problem, both from theoretical and practical perspectives. We build two kinds of GNN models to learn the solution of MaxSAT instances from benchmarks, and show that GNNs have attractive potential to solve MaxSAT problem through experimental evaluation. We also present a theoretical explanation of the effect that GNNs can learn to solve MaxSAT problem to some extent for the first time, based on the algorithmic alignment theory.

show abstract

Section: Related Workmentioning

confidence: 99%

Can Graph Neural Networks Learn to Solve MaxSAT Problem?

Liu¹,

Jia²,

Huang³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…several works [13] try to use reinforcement learning to generate item sets and avoid enumerating exponential possible combinations. Like RL-based RS, current NCO works [11,2] also lack attention to real datasets, offline policy evaluation, and extrapolation error. We expect RL4RS resources can also contribute to research in neural combinatorial optimization.…”

Section: Related Researchmentioning

confidence: 99%

RL4RS: A Real-World Benchmark for Reinforcement Learning based Recommender System

Wang¹,

Zou²,

Zhao³

et al. 2021

Preprint

View full text Add to dashboard Cite

Reinforcement learning based recommender systems (RL-based RS) aims at learning a good policy from a batch of collected data, with casting sequential recommendation to multi-step decisionmaking tasks. However, current RL-based RS benchmarks commonly have a large reality gap, because they involve artificial RL datasets or semi-simulated RS datasets, and the trained policy is directly evaluated in the simulation environment. In realworld situations, not all recommendation problems are suitable to be transformed into reinforcement learning problems. Unlike previous academic RL researches, RL-based RS suffer from extrapolation error and the difficulties of being well validated before deployment.In this paper, we introduce the RL4RS (Reinforcement Learning for Recommender Systems) benchmark -a new resource fully collected from industrial applications to train and evaluate RL algorithms with special concerns on the above issues. It contains two datasets, tuned simulation environments, related advanced RL baselines,data understanding tools, and counterfactual policy evaluation algorithms. The RL4RS suit can be found at https://github.com/fuxiAIlab/RL4RS. In addition to the RLbased recommender systems, we expect the resource to contribute to research in reinforcement learning and neural combinatorial optimization.

show abstract

“…Graph Neural Network (gnn-hist): In this model, we employ the encoder-decoder architecture used in many combinatorial optimization problems, see [6]. At each timestep t, the graph encoder consumes the current graph and produces embeddings for all nodes.…”

Section: Deep Learning Architecturesmentioning

confidence: 99%

Deep Policies for Online Bipartite Matching: A Reinforcement Learning Approach

Ali¹,

Reza²,

Khalil³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

From assigning computing tasks to servers and advertisements to users, sequential online matching problems arise in a wide variety of domains. The challenge in online matching lies in making irrevocable assignments while there is uncertainty about future inputs. In the theoretical computer science literature, most policies are myopic or greedy in nature. In real-world applications where the matching process is repeated on a regular basis, the underlying data distribution can be leveraged for better decision-making. We present an end-to-end Reinforcement Learning framework for deriving better matching policies based on trial-and-error on historical data. We devise a set of neural network architectures, design feature representations, and empirically evaluate them across two online matching problems: Edge-Weighted Online Bipartite Matching and Online Submodular Bipartite Matching. We show that most of the learning approaches perform significantly better than classical greedy algorithms on four synthetic and real-world datasets. Our code is publicly available at https://github.com/lyeskhalil/CORL.git.

show abstract

Combinatorial Optimization and Reasoning with Graph Neural Networks

Cited by 83 publications

References 3 publications

Can Graph Neural Networks Learn to Solve MaxSAT Problem?

Can Graph Neural Networks Learn to Solve MaxSAT Problem?

RL4RS: A Real-World Benchmark for Reinforcement Learning based Recommender System

Deep Policies for Online Bipartite Matching: A Reinforcement Learning Approach

Contact Info

Product

Resources

About