Article
Physics, Multidisciplinary
Baki Unal
Summary: This study analyzes the causalities of COVID-19 among seventy countries using effective transfer entropy. It constructs a weighted directed network to reveal the strength of the causality, which is obtained by calculating effective transfer entropies. Transfer entropy has the advantage of quantifying the strength of causality and detecting nonlinear causal relationships. The causality network is then analyzed using network analysis methods such as eigenvector centrality, PageRank, and community detection. Eigenvector centrality and PageRank metrics reveal the importance and centrality of each country in the network, while community detection groups node countries with denser connections.
Article
Chemistry, Multidisciplinary
Xiaomiao Song, Qinglong Liu, Mingxin Dong, Yifei Meng, Chuanrui Qin, Dongfeng Zhao, Fabo Yin, Jiangbo Jiu
Summary: This study combines knowledge-driven and data-driven methods to establish an alarm causal network model for root cause diagnosis of alarms, successfully applied in a real petrochemical plant, improving accuracy and reducing misjudgment probability of operators.
Article
Computer Science, Information Systems
Chainarong Amornbunchornvej, Elena Zheleva, Tanya Berger-Wolf
Summary: Granger causality and Transfer Entropy are commonly used in causal inference in time series data, but they often assume fixed time delays. To address this issue, Variable-lag Granger causality and Variable-lag Transfer Entropy have been developed to allow causes to influence effects with arbitrary time delays.
ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA
(2021)
Article
Mathematics, Applied
Yujia Mi, Aijing Lin
Summary: Transfer entropy is used to quantify information flow and causal orientation in nonlinear systems. This study combines noise-assisted multivariate variational mode decomposition (NA-MVMD) and transfer entropy to propose a kernel-based multiscale partial Renyi transfer entropy for multivariate systems. The method is validated using henon mapping, VAR model, and multi-channel EEG signals, demonstrating its robustness to noise and its ability to measure information transfer at different scales.
COMMUNICATIONS IN NONLINEAR SCIENCE AND NUMERICAL SIMULATION
(2023)
Article
Physics, Multidisciplinary
Adrian Moldovan, Angel Cataron, Razvan Andonie
Summary: The article discusses the integration of Transfer Entropy (TE) feedback connections in a Convolutional Neural Network (CNN) architecture to accelerate training process and improve stability by considering TE between neuron pairs in the last two fully connected layers.
Article
Physics, Multidisciplinary
Semei Coronado, Jose N. Martinez, Victor Gualajara, Omar Rojas
Summary: The relationship between COVID-19 news series and stock market volatility in Latin American countries and the U.S. was analyzed. The results showed that the U.S. and Latin American stock markets reacted differently to COVID-19 news, and certain news indices had significant impact on Latin American markets. These COVID-19 news indices can be used to forecast stock market volatility in the U.S. and Latin America.
Article
Physics, Multidisciplinary
Hongduo Cao, Fan Lin, Ying Li, Yiming Wu
Summary: The study investigates how price fluctuations of a sovereign currency are transmitted among currencies and the network traits formed in this process under the background of economic globalization. It was found that there may be strong information exchange between currencies when the overall market price fluctuates violently. Commodity currencies and currencies of major countries have great influence in the network, and local fluctuations may result in increased risks in the overall exchange rate market.
Article
Engineering, Mechanical
Pengfei Wang, Yixuan Guo, Zhenkun Xu, Weihao Wang, Diyi Chen
Summary: The hydropower generation system is a complex nonlinear system with hybrid state responses. This study proposes a data mining and data prediction strategy based on information causality and the PageRank algorithm, which is validated using a 250 MW hydropower unit. The results show that considering the information transfer sequences between variables improves prediction accuracy.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Mathematics, Applied
Peter Jan van Leeuwen, Michael DeCaria, Nachiketa Chakraborty, Manuel Pulido
Summary: The study introduces a new framework for inferring causal relationships in complex nonlinear systems, which provides complete information theoretic disentanglement and handles nonlinear causal interactions. The framework is built upon information theoretic measures that gradually increase the information available about the target process. Additionally, it can analyze systems that cannot be represented on directed acyclic graphs.
Article
Mathematics, Applied
Chun-Xiao Nie
Summary: This paper utilizes transfer entropy and surrogates to examine the information flow between price and transaction volume, revealing stronger information flow during stock bubble bursts or financial crises and presenting a new approach to analyzing the price-volume relationship.
Article
Economics
Maximo Camacho, Andres Romeu, Manuel Ruiz-Marin
Summary: The study introduces a non-parametric Granger causality test procedure for longitudinal data using multiple-unit symbolic dynamics and transfer entropy. Monte Carlo simulations demonstrate that the test maintains correct size and high power in situations where linear panel data causality tests do not work. The usefulness of the proposed procedure is illustrated through dynamic causal relationship analysis in various economic contexts.
ECONOMIC MODELLING
(2021)
Article
Multidisciplinary Sciences
Vittoria Volta, Tomaso Aste
Summary: This study investigates the high-frequency reactions in the Eurozone and UK stock markets during the time period surrounding the interest rate decisions of the ECB and BoE. The effects are assessed by quantifying linear and nonlinear transfer entropy, combined with bivariate empirical mode decomposition. The findings reveal that central bank interest rate decisions lead to an increase in intraday volatility, particularly on ECB announcement days, and there is significant information flow between the two markets, predominantly from the market where the announcement is made to the other market.
ROYAL SOCIETY OPEN SCIENCE
(2022)
Article
Multidisciplinary Sciences
Manuel Castro, Pedro Ribeiro Mendes Junior, Aurea Soriano-Vargas, Rafael de Oliveira Werneck, Maiara Moreira Goncalves, Leopoldo Lusquino Filho, Renato Moura, Marcelo Zampieri, Oscar Linares, Vitor Ferreira, Alexandre Ferreira, Alessandra Davolio, Denis Schiozer, Anderson Rocha
Summary: In this study, we propose using ensemble models (such as Random Forest) to assess the importance of input features in machine learning models, in order to establish causal relationships between variables. By analyzing oil field production data, we find that our results align with confirmed tracer information, demonstrating the effectiveness of our proposed methodology.
SCIENTIFIC REPORTS
(2023)
Article
Automation & Control Systems
Xiangxiang Zhang, Wenkai Hu, Fan Yang, Weihua Cao, Min Wu
Summary: This paper proposes a new transfer entropy approach based on information granulation and clustering to identify the root causes of faults in complex industrial facilities. The approach includes information granulation based transfer entropy and information granulation based direct transfer entropy, as well as a PDF estimator based on OPTICS clustering. The effectiveness of the proposed approach is demonstrated through two case studies.
CONTROL ENGINEERING PRACTICE
(2023)
Article
Computer Science, Artificial Intelligence
Junya Chen, Jianfeng Feng, Wenlian Lu
Summary: The study focuses on the fundamental task of discovering causal relationships in investigating the dynamics of complex systems. A novel definition of Wiener causality based on relative entropy is proposed, and it is argued that any Bregman divergences can be used for detecting causal relations. The discussion includes the benefits of different choices of divergence functions on causal inference and the quality of obtained causal models, with experimental evidence provided on how these causalities improve detection accuracy.
NEURAL PROCESSING LETTERS
(2021)
Article
Biochemistry & Molecular Biology
Christopher R. S. Banerji, Maryna Panamarova, Johanna Pruller, Nicolas Figeac, Husam Hebaishi, Efthymios Fidanis, Alka Saxena, Julian Contet, Sabrina Sacconi, Simone Severini, Peter S. Zammit
HUMAN MOLECULAR GENETICS
(2019)
Article
Biochemistry & Molecular Biology
Christopher R. S. Banerji, Peter S. Zammit
HUMAN MOLECULAR GENETICS
(2019)
Article
Cell Biology
Nicolas Figeac, Johanna Pruller, Isabella Hofer, Mathieu Fortier, Huascar Pedro Ortuste Quiroga, Christopher R. S. Banerji, Peter S. Zammit
CELL PROLIFERATION
(2020)
Article
Mathematics, Applied
M. Lupini, L. Mancinska, V. I. Paulsen, D. E. Roberson, G. Scarpa, S. Severini, I. G. Todorov, A. Winter
MATHEMATICAL PHYSICS ANALYSIS AND GEOMETRY
(2020)
Article
Biochemistry & Molecular Biology
Christopher R. S. Banerji, Maryna Panamarova, Peter S. Zammit
HUMAN MOLECULAR GENETICS
(2020)
Article
Clinical Neurology
Christopher R. S. Banerji, Phillip Cammish, Teresinha Evangelista, Peter S. Zammit, Volker Straub, Chiara Marini-Bettolo
NEUROMUSCULAR DISORDERS
(2020)
Article
Biochemistry & Molecular Biology
Christopher R. S. Banerji, Don Henderson, Rabi N. Tawil, Peter S. Zammit
HUMAN MOLECULAR GENETICS
(2020)
Article
Mathematics, Applied
Joshua Lockhart, Simone Severini
Summary: The paper introduces new combinatorial objects, grid-labelled graphs, to represent quantum states arising in a specific physical scenario. By reformulating entanglement criteria, new bound entangled states are constructed and the limitations of matrix realignment are demonstrated. The relationship between local operations and classical communication (LOCC) and a generalisation of the graph isomorphism problem is also discussed.
LINEAR ALGEBRA AND ITS APPLICATIONS
(2021)
Review
Medicine, Research & Experimental
Christopher R. S. Banerji, Peter S. Zammit
Summary: Facioscapulohumeral muscular dystrophy (FSHD) is characterized by skeletal muscle weakness and wasting due to epigenetic derepression of the D4Z4 macrosatellite, leading to transcription of DUX4 which activates target genes. PAX7 suppression serves as a reliable biomarker for FSHD, but its link to genomic changes and DUX4 remains unclear. Understanding the roles of DUX4 and PAX7 in FSHD pathology can deepen knowledge of the disease through interactions with the immune system and muscle regeneration.
EMBO MOLECULAR MEDICINE
(2021)
Article
Biochemistry & Molecular Biology
Philipp Heher, Massimo Ganassi, Adelheid Weidinger, Elise N. Engquist, Johanna Pruller, Thuy Hang Nguyen, Alexandra Tassin, Anne-Emilie Decleves, Kamel Mamchaoui, Christopher R. S. Banerji, Johannes Grillari, Andrey V. Kozlov, Peter S. Zammit
Summary: FSHD is characterized by oxidative stress induced by DUX4, leading to metabolic dysfunction and impaired mitochondrial function. Increased mitochondrial ROS levels in FSHD muscle cells are associated with elevated steady-state mitochondrial membrane potential. DUX4 triggers mitochondrial membrane polarization, resulting in mitochondrial ROS generation and apoptosis.
Editorial Material
Biochemistry & Molecular Biology
Christopher R. S. Banerji, Tapabrata Chakraborti, Chris Harbron, Ben D. Macarthur
Summary: Personalized measures of uncertainty, utilizing techniques like conformal prediction, are crucial for clinical artificial intelligence to reach its potential and enhance human health.
Article
Physics, Multidisciplinary
Andrew Patterson, Hongxiang Chen, Leonard Wossnig, Simone Severini, Dan Browne, Ivan Rungger
Summary: In the near term, noisy quantum computers require algorithms with low circuit depth and qubit count. Research shows that introducing a smaller circuit ansatz can overcome the limitations of gradient calculation on noisy devices with a large number of parameters. The main effect of noise is to increase the overlap between states as circuit gates are applied, making discrimination more challenging.
PHYSICAL REVIEW RESEARCH
(2021)
Article
Quantum Science & Technology
Alessandro Rudi, Leonard Wossnig, Carlo Ciliberto, Andrea Rocchetto, Massimiliano Pontil, Simone Severini
Article
Computer Science, Theory & Methods
Varun Kanade, Andrea Rocchetto, Simone Severini
QUANTUM INFORMATION & COMPUTATION
(2019)
Proceedings Paper
Computer Science, Information Systems
Abdulah Fawaz, Paul Klein, Sebastien Piat, Simone Severini, Peter Mountney
KDD'19: PROCEEDINGS OF THE 25TH ACM SIGKDD INTERNATIONAL CONFERENCCE ON KNOWLEDGE DISCOVERY AND DATA MINING
(2019)