Using Machine Learning to Assess Short Term Causal Dependence and Infer Network Links

@article{Banerjee2019UsingML,
  title={Using Machine Learning to Assess Short Term Causal Dependence and Infer Network Links},
  author={Amitava Banerjee and Jaideep Pathak and Rajarshi Roy and Juan G. Restrepo and Edward Ott},
  journal={Chaos},
  year={2019},
  volume={29 12},
  pages={
          121104
        }
}
We introduce and test a general machine-learning-based technique for the inference of short term causal dependence between state variables of an unknown dynamical system from time-series measurements of its state variables. Our technique leverages the results of a machine learning process for short time prediction to achieve our goal. The basic idea is to use the machine learning to estimate the elements of the Jacobian matrix of the dynamical flow along an orbit. The type of machine learning… 

Figures and Tables from this paper

Link inference of noisy delay-coupled networks: Machine learning and opto-electronic experimental tests
TLDR
A machine learning technique to solve the general problem of inferring network links that have time-delays and is motivated by the widely-used invasive network inference method whereby the responses to active perturbations applied to the network are observed and employed to infer network links.
Multifunctionality in a Reservoir Computer
TLDR
This paper explores the critical effects that changes in certain parameters can have on the reservoir computers' ability to express multifunctionality and exposes the existence of several "untrained attractors"; attractors that dwell within the prediction state space of the reservoir computer were not part of the training.
Iterative procedure for network inference
Reducing echo state network size with controllability matrices
TLDR
A controllability matrix is formed whose rank indicates the active sub-manifold and candidate replacement reservoir size and results in time speed-ups and reduced memory usage with minimal error increase to chaotic climate reconstruction or short term prediction.
Backpropagation on Dynamical Networks
TLDR
A network inference method based on the backpropagation through time (BPTT) algorithm commonly used to train recurrent neural networks that was found to be succesful in identifying the connectivity structure for Lorenz, Chua and FitzHugh-Nagumo oscillators.
A new approach for extracting information from protein dynamics
TLDR
This study applies the inverse approach to the dynamics of protein dihedral angles, a system of internal coordinates, to avoid structural alignment in order to extract a network from protein dynamics.
Role of assortativity in predicting burst synchronization using echo state network.
TLDR
This paper presents a meta-analyses of the determinants of earthquake-triggered landsliding in central Asia and its consequences on the strength of the number of moving parts and the intensity of the earthquake.
Parallel Machine Learning for Forecasting the Dynamics of Complex Networks
TLDR
The utility and scalability of the method implemented using reservoir computing on a chaotic network of oscillators, using a parallel architecture that mimics the topology of the network of interest, are demonstrated.
Machine learning evaluates changes in functional connectivity under a prolonged cognitive load.
TLDR
This study considers the alterations in the cortical synchrony under a prolonged cognitive load and highlights the advances of this machine learning method in building a robust classification model and percept-related prestimulus connectivity changes over the conventional trial-averaged statistical analysis.
Optimized ensemble deep learning framework for scalable forecasting of dynamics containing extreme events
TLDR
This study develops an optimized ensemble deep learning framework wherein these two machine learning techniques are jointly used to achieve synergistic improvements in model accuracy, stability, scalability, and reproducibility, prompting a new wave of applications in the forecasting of dynamics.
...
...

References

SHOWING 1-10 OF 51 REFERENCES
Reconstructing dynamical networks via feature ranking
TLDR
This paper designs a new method of network reconstruction that essentially makes no assumptions about the properties and dynamics of the scrutinized network and uses two independent feature ranking approaches-Random Forest and RReliefF-to rank the importance of each node for predicting the value of each other node, which yields the reconstructed adjacency matrix.
Inferring directed networks using a rank-based connectivity measure.
TLDR
This work uses a rank-based nonlinear interdependence measure originally developed for pairs of signals to measure the strength but also the direction of the coupling, and shows an application to multichannel electroencephalographic recordings from an epilepsy patient.
Reservoir observers: Model-free inference of unmeasured variables in chaotic systems.
TLDR
It is shown that the reservoir observer can be a very effective and versatile tool for robustly reconstructing unmeasured dynamical system variables.
Reconstructing directional causal networks with random forest: Causality meeting machine learning.
TLDR
The ICA framework is designed in a network level and fills the gap between traditional mutual causality detection methods and the reconstruction of causal networks.
Granger Causality: Basic Theory and Application to Neuroscience
TLDR
The main goal of this article is to provide an expository introduction to the concept of Granger causality and the results are shown to be physiologically interpretable and yield new insights into the dynamical organization of large-scale oscillatory cortical networks.
Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations
TLDR
A new computational model for real-time computing on time-varying input that provides an alternative to paradigms based on Turing machines or attractor neural networks, based on principles of high-dimensional dynamical systems in combination with statistical learning theory and can be implemented on generic evolved or found recurrent circuitry.
Inferring network topology from complex dynamics
TLDR
This work presents an analytical solution to the inverse problem of finding the network topology from observing a time series of state variables only, and provides a conceptually new step towards reconstructing a variety of real-world networks, including gene and protein interaction networks and neuronal circuits.
Model reconstruction from temporal data for coupled oscillator networks.
TLDR
It is demonstrated that, given sufficient observational data on the transient evolution of each oscillator, machine learning can reconstruct the interaction network and identify the intrinsic dynamics.
...
...