Article
Automation & Control Systems
Randa Herzallah, David Lowe, Yazan Qarout
Summary: There is currently a lack of comprehensive understanding on how to apply fully decentralized control to networks of sparsely coupled nonlinear dynamical subsystems subject to noise to track a desired state. Research results demonstrate that utilizing probability theory and local message passing can implicitly infer global knowledge, enabling control of a global state through decentralized control signals applied to local subsystems without reference to a global current state.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2022)
Article
Automation & Control Systems
Yuyang Zhou, Randa Herzallah
Summary: This paper proposes a general decentralised probabilistic control framework for complex stochastic systems, utilizing probabilistic models to describe subsystem dynamics and using Mixture Density Networks and probabilistic message passing for control.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2021)
Article
Computer Science, Artificial Intelligence
Semih Akbayrak, Ismail Senoz, Alp Sari, Bert de Vries
Summary: Stochastic approximation methods for variational inference have gained popularity in probabilistic programming for their automation, online scalability, and universal approximate Bayesian inference. However, current Probabilistic Programming Languages (PPLs) with stochastic approximation engines lack the efficiency of message passing algorithms with deterministic update rules. This paper casts Stochastic Variational Inference (SVI) and Conjugate-Computation Variational Inference (CVI) explicitly in a message passing context, providing an implementation in ForneyLab that extends the automated inference capabilities of message passing-based probabilistic programming.
INTERNATIONAL JOURNAL OF APPROXIMATE REASONING
(2022)
Article
Multidisciplinary Sciences
M. E. J. Newman
Summary: Networks and network computations have become essential for analyzing complex systems. Message passing methods, which involve the propagation of information between network nodes, are commonly used for calculating quantities on nodes. This perspective article discusses the application of message passing methods, provides examples and applications, and explores the connection between message passing and phase transitions in networks. It also discusses the limitations of message passing methods and describes recent methods that address these limitations.
PROCEEDINGS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND ENGINEERING SCIENCES
(2023)
Article
Engineering, Electrical & Electronic
Randa Herzallah, Yuyang Zhou
Summary: This paper proposes a unified probabilistic decentralised control and message passing framework for real-time control of the electrical grid, enabling the development of future smart grids. The key elements include the design of local randomised controllers and probabilistic message passing methodology for optimal system operation coordination. Simulation studies demonstrate the applicability and effectiveness of the proposed approach in a multi-area power system.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2021)
Article
Mechanics
Marco Mondelli, Ramji Venkataramanan
Summary: The study focuses on estimating signals from measurements obtained via a generalized linear model, proposing an AMP algorithm initialized with a spectral estimator and rigorously characterizing its performance in the high-dimensional limit.
JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT
(2022)
Article
Mechanics
Burak Cakmak, Yue M. Lu, Manfred Opper
Summary: This paper analyzes a random sequential message passing algorithm for large Gaussian latent variable models. By assuming random covariance matrices and considering model mismatch, the authors obtain dynamical mean-field equations characterizing the dynamics of the inference algorithm, and derive the parameter range for which the algorithm does not converge.
JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT
(2022)
Article
Computer Science, Artificial Intelligence
Zonghan Wu, Da Zheng, Shirui Pan, Quan Gan, Guodong Long, George Karypis
Summary: This article introduces a novel spatial-temporal graph neural network called TraverseNet for capturing the spatial-temporal dependencies in traffic data. Compared to other spatial-temporal neural networks, TraverseNet views space and time as an inseparable whole and utilizes message traverse mechanisms to explore the dependencies in the spatial-temporal graph.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Mechanics
Giovanni Piccioli, Guilhem Semerjian, Gabriele Sicuro, Lenka Zdeborova
Summary: The study investigates a polynomial time message-passing algorithm designed to solve the inference problem of partially recovering the hidden permutation, in the sparse regime with constant average degrees. Extensive numerical simulations are conducted to determine the range of parameters in which this algorithm achieves partial recovery, and the algorithm is also extended to a generalized ensemble of correlated random graphs with prescribed degree distributions.
JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT
(2022)
Article
Acoustics
Xisheng Wu, Dong Li, Yanbo Wu, Min Zhu
Summary: This study proposes a probabilistic constellation shaping (PCS)-aided single-carrier transceiver to enhance the spectral efficiency of underwater acoustic (UWA) communications. The PCS is realized by mapping coded bits onto a quadrature amplitude modulation constellation. An improved frequency-domain turbo equalizer based on the vector approximate message passing (VAMP) is used to eliminate multipath interference. Experimental results demonstrate the superiority of the PCS-UWA communication system compared to traditional systems and the proposed receiver outperforms the classical adaptive turbo equalizer.
JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA
(2023)
Article
Physics, Fluids & Plasmas
Giuseppe Torrisi, Alessia Annibale, Reimer Kuhn
Summary: The dynamic cavity method is efficient for evaluating probabilities of dynamic trajectories in systems of stochastic units with unidirectional sparse interactions, but faces challenges related to complexity barriers associated with in-degrees.
Article
Engineering, Electrical & Electronic
Dan Zhang, Xiaohang Song, Wenjin Wang, Gerhard Fettweis, Xiqi Gao
Summary: This article unifies variational message passing, belief propagation, and expectation propagation under an optimization framework of Bethe free energy minimization with differently imposed constraints, providing a theoretical framework for systematically deriving message passing variants. By reformulating constraints, a low-complexity EP variant is obtained for better estimation performance. Furthermore, a hybrid message passing algorithm is systematically derived for joint SSR and statistical model learning with near-optimal inference performance and scalable complexity.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2021)
Article
Astronomy & Astrophysics
A. Bhat, D. Malyshev
Summary: This article discusses the probabilistic classification of Fermi-LAT sources using machine learning methods. The authors determine the classification of pulsars, active galactic nuclei (AGNs), and other sources by comparing different meta-parameters of the machine learning methods. The results show that the three-class classification performs similarly to the two-class classification in terms of reliability and does not require adjustment for the presence of other sources.
ASTRONOMY & ASTROPHYSICS
(2022)
Article
Multidisciplinary Sciences
Jakub Rybak, Heather S. Battey
Summary: This paper explores embeddings for relevant covariance models to be sparse, utilizing skew-symmetric matrices for parameterization and exploration of sparsity.
PROCEEDINGS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND ENGINEERING SCIENCES
(2021)
Article
Computer Science, Information Systems
Takashi Takahashi, Yoshiyuki Kabashima
Summary: In this study, the macroscopic properties of the VAMP algorithm for inference of generalized linear models are investigated using a non-rigorous heuristic method of statistical mechanics. The study focuses on the correspondence between the non-rigorous replica analysis of statistical mechanics and the performance assessment of VAMP in the model-mismatched setting. The results show that the fixed point of VAMP is generally consistent with the replica symmetric solution obtained through the replica method of statistical mechanics. The study also reveals that the fixed point of VAMP can exhibit a microscopic instability.
IEEE TRANSACTIONS ON INFORMATION THEORY
(2022)
Article
Chemistry, Analytical
Yazan Qarout, Yordan P. Raykov, Max A. Little
Article
Health Care Sciences & Services
Luc J. W. Evers, Yordan P. Raykov, Jesse H. Krijthe, Ana Ligia Silva de Lima, Reham Badawy, Kasper Claes, Tom M. Heskes, Max A. Little, Marjan J. Meinders, Bastiaan R. Bloem
JOURNAL OF MEDICAL INTERNET RESEARCH
(2020)
Article
Physics, Multidisciplinary
Alexander Mozeika, Bo Li, David Saad
PHYSICAL REVIEW LETTERS
(2020)
Article
Physics, Multidisciplinary
Mihai-Alin Badiu, David Saad, Justin P. Coon
Summary: The proposed self-organization scheme for routing in multi-hop networks balances route costs and node loads by penalizing high loads and applying belief propagation. Through numerical demonstration, the framework's efficacy in balancing node loads is shown, and the trade-off between load reduction and total cost minimization is studied.
JOURNAL OF PHYSICS A-MATHEMATICAL AND THEORETICAL
(2021)
Article
Acoustics
Amir Hossein Poorjam, Mathew Shaji Kavalekalam, Liming Shi, Jordan P. Raykov, Jesper Rindom Jensen, Max A. Little, Mads Graesboll Christensen
Summary: This study investigates the impact of various acoustic degradations on the performance of voice-based Parkinson's disease detection systems, and proposes two methods for automatically controlling the quality of recordings to improve PD detection accuracy. Experimental results demonstrate the effectiveness of quality control approaches in selecting appropriate enhancement methods, leading to improved PD detection accuracy.
SPEECH COMMUNICATION
(2021)
Article
Physics, Multidisciplinary
Hanlin Sun, David Saad, Andrey Y. Lokhov
Summary: Competition and collaboration play crucial roles in multiagent probabilistic spreading processes, with examples including competitive marketing campaigns and joint spread of infectious diseases. By deriving dynamic message-passing equations and developing low-complexity models, the dynamics of spreading processes on networks can be predicted. A theoretical framework for optimal control through optimized resource allocation has been proposed, and the efficacy of the framework and optimization method has been demonstrated on both synthetic and real-world networks.
Article
Computer Science, Information Systems
Yordan P. Raykov, Luc J. W. Evers, Reham Badawy, Bastiaan R. Bloem, Tom M. Heskes, Marjan J. Meinders, Kasper Claes, Max A. Little
Summary: In this study, a principled modeling approach for free-living gait analysis was developed to support health predictions and clinical diagnosis. Using a dataset of PD patients and controls, the framework's effectiveness in detecting gait and predicting medication-induced fluctuations in PD patients was demonstrated. The approach was shown to be robust to varying sensor locations.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2021)
Article
Physics, Multidisciplinary
Yi-Zhi Xu, David Saad
Summary: This paper presents a message passing-based framework for analyzing and addressing the optimization problem of edge removal/addition in networks. The methods developed result in improved path lengths and higher network capacity.
PHYSICAL REVIEW RESEARCH
(2023)
Article
Physics, Fluids & Plasmas
Bo Li, David Saad, Chi Ho Yeung
Summary: Optimizing embedded systems is a formidable challenge that is ubiquitous in real world systems. This study focuses on flow networks and employs message passing algorithms to optimize a global objective.
Article
Physics, Fluids & Plasmas
Yi-Zhi Xu, Ho Fai Po, Chi Ho Yeung, David Saad
Summary: Probabilistic message-passing algorithms are developed to address the routing transmissions problem in multiwavelength optical communication networks. The method provides good approximate solutions on locally treelike graphs and accommodates various objective functions. It can be used for managing and designing optical communication networks and settles the debate on the merit of wavelength-switching nodes.
Article
Physics, Fluids & Plasmas
Bo Li, David Saad
Summary: The study focuses on the variant model of infectious diseases with presymptomatic transmission, using the method of dynamic message passing to provide a good estimate of the probabilistic evolution of spread. This facilitates the derivation of epidemic thresholds and impacts different containment strategies.
Article
Physics, Fluids & Plasmas
Ho Fai Po, Chi Ho Yeung, David Saad
Summary: The study shows that a small ratio of selfish route choices can improve the overall performance of uncoordinated transportation networks while degrading the efficiency of optimized systems. Compliant users always benefit in this scenario, and selfish users may also gain under specific conditions. Iterative route switching by a small fraction of selfish users can lead to Nash equilibria close to the globally optimal routing solution.
Article
Physics, Multidisciplinary
Bo Li, David Saad, Andrey Y. Lokhov
PHYSICAL REVIEW RESEARCH
(2020)