9 research outputs found
Parameterizing Markov Networks using Graph Neural Networks for Trackster Pruning at the CERN CMS High-Granularity Calorimeter
No full text
Node Classification in Random Trees
Full text linkWe propose a method for the classification of objects that are structured as random trees. Our aim is to model a distribution over the node label assignments in settings where the tree data structure is associated with node attributes (typically high dimensional embeddings). The tree topology is not predetermined and none of the label assignments are present during inference. Other methods that produce a distribution over node label assignment in trees (or more generally in graphs) either assume conditional independence of the label assignment, operate on a fixed graph topology, or require part of the node labels to be observed. Our method defines a Markov Network with the corresponding topology of the random tree and an associated Gibbs distribution. We parameterize the Gibbs distribution with a Graph Neural Network that operates on the random tree and the node embeddings. This allows us to estimate the likelihood of node assignments for a given random tree and use MCMC to sample from the distribution of node assignments. We evaluate our method on the tasks of node classification in trees on the Stanford Sentiment Treebank dataset. Our method outperforms thebaselines on this dataset, demonstrating its effectiveness for modeling joint distributions of node labels in random trees
Long-term planning of large interventions within complex and dynamic infrastructure systems: Introducing a decision-support method for strategic intervention planning
No full textCities are rapidly transforming, and infrastructures are not always resistant to these future developments. On a global scale cities are growing into mega cities, whilst counter urbanization causes significant population declines. To adapt infrastructures against these future developments, infrastructure asset management can be applied. Infrastructure asset managers require reliable insight in future developments, to make their intervention decisions strategically. This involves assessing multiple variables, and their underlying relations. Infrastructure systems can perform multiple functions, have various connecting interfaces, and are mostly situated in a dynamically changing environment. Therefore, intervention decisions are subjected to dynamic complexity, and uncertainty. However, asset managers predominantly make large intervention decisions based on decision support methods of a static nature, which provide insufficient insight in the dynamic complexity, and uncertainty, of future developments. Since infrastructures are the backbones of local economies, a decision support method able to incorporate all relevant complexity, and uncertainties, is inadmissible for infrastructure asset management. This study aims at improving large intervention decisions, by assessing dynamic complexity, and uncertainty, with multivariate simulation approaches as the exploratory system dynamics modelling and analysis approach (ESDMA), and adapting strategies for large intervention decisions with the adaptation pathways approach. The approaches were applied to a case study, which is a highly schematized representation of the city of Amsterdam, with its interconnected infrastructure network. The study showed that the proposed approaches can improve large interventions decisions on an infrastructure network level, by adapting them dynamically over time to uncertainty, and complexity. The key findings include the identification of opportunities, ‘no regret actions’, dependencies, effects on interconnected infrastructure systems, and the required intervention timing for multi-dimensional functions. A most adaptive set of interventions for all future scenarios could be identified on the basis of costs, target effects, and possible side effects. Infrastructure System Dynamics (INDY)Civil Engineering | Construction Management and Engineerin
Reactive Environments for Active Inference Agents with RxEnvironments.jl
No full textActive Inference is a framework that emphasizes the interaction between agents and their environment. While the framework has seen significant advancements in the development of agents, the environmental models are often borrowed from reinforcement learning problems, which may not fully capture the complexity of multi-agent interactions or allow complex, conditional communication. This paper introduces Reactive Environments, a comprehensive paradigm that facilitates complex multi-agent communication. In this paradigm, both agents and environments are defined as entities encapsulated by boundaries with interfaces. This setup facilitates a robust framework for communication in nonequilibrium-Steady-State systems, allowing for complex interactions and information exchange. We present a Julia package RxEnvironments.jl, which is a specific implementation of Reactive Environments, where we utilize a Reactive Programming style for efficient implementation. The flexibility of this paradigm is demonstrated through its application to several complex, multi-agent environments. These case studies highlight the potential of Reactive Environments in modeling sophisticated systems of interacting agents
GraphPPL.jl: A Probabilistic Programming Language for Graphical Models
Full text linkThis paper presents GraphPPL.jl, a novel probabilistic programming language designed for graphical models. GraphPPL.jl uniquely represents probabilistic models as factor graphs. A notable feature of GraphPPL.jl is its model nesting capability, which facilitates the creation of modular graphical models and significantly simplifies the development of large (hierarchical) graphical models. Furthermore, GraphPPL.jl offers a plugin system to incorporate inference-specific information into the graph, allowing integration with various well-known inference engines. To demonstrate this, GraphPPL.jl includes a flexible plugin to define a Constrained Bethe Free Energy minimization process, also known as variational inference. In particular, the Constrained Bethe Free Energy defined by GraphPPL.jl serves as a potential inference framework for numerous well-known inference backends, making it a versatile tool for diverse applications. This paper details the design and implementation of GraphPPL.jl, highlighting its power, expressiveness, and user-friendliness. It also emphasizes the clear separation between model definition and inference while providing developers with extensibility and customization options. This establishes GraphPPL.jl as a high-level user interface language that allows users to create complex graphical models without being burdened with the complexity of inference while allowing backend developers to easily adopt GraphPPL.jl as their frontend language
Online Structure Learning with Dirichlet Processes Through Message Passing
No full textGenerative or probabilistic modeling is crucial for developing intelligent agents that can reason about their environment. However, designing these models manually for complex tasks is often infeasible. Structure learning addresses this challenge by automating model creation based on sensory observations, balancing accuracy with complexity. Central to structure learning is Bayesian model comparison, which provides a principled framework for evaluating models based on their evidence. This paper focuses on model expansion and introduces an online message passing procedure using Dirichlet processes, a prominent prior in non-parametric Bayesian methods. Our approach builds on previous work by automating Bayesian model comparison using message passing based on variational free energy minimization. We derive novel message passing update rules to emulate Dirichlet processes, offering a flexible and scalable method for online structure learning. Our method generalizes to arbitrary models and treats structure learning identically to state estimation and parameter learning. The experimental results validate the effectiveness of our approach on an infinite mixture model
Automating Model Comparison in Factor Graphs
Full text linkBayesian state and parameter estimation have been automated effectively in a
variety of probabilistic programming languages. The process of model comparison
on the other hand, which still requires error-prone and time-consuming manual
derivations, is often overlooked despite its importance. This paper efficiently
automates Bayesian model averaging, selection, and combination by message
passing on a Forney-style factor graph with a custom mixture node. Parameter
and state inference, and model comparison can then be executed simultaneously
using message passing with scale factors. This approach shortens the model
design cycle and allows for the straightforward extension to hierarchical and
temporal model priors to accommodate for modeling complicated time-varying
processes
Correction: The Influence of Age and Sex on Genetic Associations with Adult Body Size and Shape: A Large-Scale Genome-Wide Interaction Study
Full text linkThe arcOGEN Consortium should be listed as an author of this article. They contributed to the genome-wide association study results presented in this work. They should be listed in the author byline at position 292 and affiliated with The Arthritis Research UK Osteoarthritis Genetics Consortium. They should also be included in the footnote designating consortia which is underneath the author affiliation list in the PDF version of the article, and in the S2 Text. Please view the correct S2 Text below, containing correct consortia members
