arXiv research

Graph neural Thompson Sampling improves online decision-making for graph data.

AI learns to design chemical processes efficiently.

Neural Architecture Search (NAS) enabled the discovery of state-of-the-art architectures in many domains. However, the success of NAS depends on the definition of the search space. Current search spaces are defined as a static sequence of decisions and a set of available actions for each decision. Each possible sequenc…

Graph Networks are used to make decisions in potentially complex scenarios but it is usually not obvious how or why they made them. In this work, we study the explainability of Graph Network decisions using two main classes of techniques, gradient-based and decomposition-based, on a toy dataset and a chemistry task. Ou…

EBNCs are a new model for Bayesian networks, derived from different networks.

Agents cooperate to make decisions in multi-armed bandits over a graph.

TD-GEN generates graphs using tree decomposition, improving efficiency and performance.

Study how communication and feedback graphs affect learning outcomes.

In this paper, we introduce a new online decision making paradigm that we call Thresholding Graph Bandits. The main goal is to efficiently identify a subset of arms in a multi-armed bandit problem whose means are above a specified threshold. While traditionally in such problems, the arms are assumed to be independent, …

Paper tackles active learning for GNNs, reducing annotation costs.

A new method for multi-agent planning on graphs outperforms existing approaches.

We consider stochastic multi-armed bandit problems with graph feedback, where the decision maker is allowed to observe the neighboring actions of the chosen action. We allow the graph structure to vary with time and consider both deterministic and Erdős-Rényi random graph models. For such a graph feedback model, we fir…

SBO uses dual voting to build consensus in noisy feedback settings.

We consider an adversarial online learning setting where a decision maker can choose an action in every stage of the game. In addition to observing the reward of the chosen action, the decision maker gets side observations on the reward he would have obtained had he chosen some of the other actions. The observation str…

We study multi-armed bandit problems with graph feedback, in which the decision maker is allowed to observe the neighboring actions of the chosen action, in a setting where the graph may vary over time and is never fully revealed to the decision maker. We show that when the feedback graphs are undirected, the original …

Graph Neural Networks improve demand forecasting by considering article relationships.

A new method for semi-supervised classification using graph walks and reinforcement learning.

Novel algorithms for online learning with uncertain feedback graphs reduce regret.

This paper extends combinatorial semi-bandits to graph feedback, improving regret bounds.

This paper presents a novel method for structural data recognition using a large number of graph models. In general, prevalent methods for structural data recognition have two shortcomings: 1) Only a single model is used to capture structural variation. 2) Naive recognition methods are used, such as the nearest neighbo…

Study of reinforcement learning with additional feedback observations.

A new framework explains GNN predictions by simulating graph structure and feature changes.

Graph-based state representation improves deep RL performance.

Molecular "fingerprints" encoding structural information are the workhorse of cheminformatics and machine learning in drug discovery applications. However, fingerprint representations necessarily emphasize particular aspects of the molecular structure while ignoring others, rather than allowing the model to make data-d…

New algorithms optimize decision-making under uncertainty with graph information.

Graph Neural Networks (GNNs) have proven to be successful in many classification tasks, outperforming previous state-of-the-art methods in terms of accuracy. However, accuracy alone is not enough for high-stakes decision making. Decision makers want to know the likelihood that a specific GNN prediction is correct. For …

We study the explore-exploit tradeoff in distributed cooperative decision-making using the context of the multiarmed bandit (MAB) problem. For the distributed cooperative MAB problem, we design the cooperative UCB algorithm that comprises two interleaved distributed processes: (i) running consensus algorithms for estim…

Interactive RL and DT feedback improve feature selection efficiency.

Method verifies robustness of graph neural networks to graph perturbations.

The paper proposes multicalibration to improve matching in graphs with imperfect predictors.

Paper formalizes a reinforcement learning model for complex information structures.

Though reinforcement learning has greatly benefited from the incorporation of neural networks, the inability to verify the correctness of such systems limits their use. Current work in explainable deep learning focuses on explaining only a single decision in terms of input features, making it unsuitable for explaining …

We present a deep reinforcement learning approach to minimizing the execution cost of neural network computation graphs in an optimizing compiler. Unlike earlier learning-based works that require training the optimizer on the same graph to be optimized, we propose a learning approach that trains an optimizer offline an…

DeepGG generates graph distributions for drug discovery and molecular design.

Proves error bounds for state representation in RL using graph spectral features.

Paper shows graphs can be embedded in lower dimensions than expected.

CI-GNN uses GNNs to diagnose psychiatric disorders by identifying causally relevant brain regions.

IDS algorithm optimizes sequential decisions in various monitoring settings.

Graph classification improved using spectral features and wavelet filters.

Improved graph embedding through refined linear transformation and community recovery.

We present an equivalent criterion for the global existence of Euler's multiplier for an integrable one-form taking into account the corresponding codim-1-foliation. In particular, the impact of inseparable leaves is considered. Here, we suppose that the foliation can be reduced to a graph; we also discuss obstructions…

GRETEL unifies GCE evaluation across various settings.

Tree Mover's Distance measures graph attributes and improves GNN performance.

We present a formal model of human decision-making in explore-exploit tasks using the context of multi-armed bandit problems, where the decision-maker must choose among multiple options with uncertain rewards. We address the standard multi-armed bandit problem, the multi-armed bandit problem with transition costs, and …

Model forecasts global stock market volatility using dynamic graphs and all trading days.

The paper introduces fixed-point centralities for networks and graphons.

We study the robustness verification problem for tree-based models, including decision trees, random forests (RFs) and gradient boosted decision trees (GBDTs). Formal robustness verification of decision tree ensembles involves finding the exact minimal adversarial perturbation or a guaranteed lower bound of it. Existin…

Graph Neural Networks (GNN) are a promising technique for bridging differential programming and combinatorial domains. GNNs employ trainable modules which can be assembled in different configurations that reflect the relational structure of each problem instance. In this paper, we show that GNNs can learn to solve, wit…