Inferring, Predicting, and Denoising Causal Wave Dynamics

09/19/2020
by   Matthias Karlbauer, et al.
0

The novel DISTributed Artificial neural Network Architecture (DISTANA) is a generative, recurrent graph convolution neural network. It implements a grid or mesh of locally parameterizable laterally connected network modules. DISTANA is specifically designed to identify the causality behind spatially distributed, non-linear dynamical processes. We show that DISTANA is very well-suited to denoise data streams, given that re-occurring patterns are observed, significantly outperforming alternative approaches, such as temporal convolution networks and ConvLSTMs, on a complex spatial wave propagation benchmark. It produces stable and accurate closed-loop predictions even over hundreds of time steps. Moreover, it is able to effectively filter noise – an ability that can be improved further by applying denoising autoencoder principles or by actively tuning latent neural state activities retrospectively. Results confirm that DISTANA is ready to model real-world spatio-temporal dynamics such as brain imaging, supply networks, water flow, or soil and weather data patterns.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
12/23/2019

A Distributed Neural Network Architecture for Robust Non-Linear Spatio-Temporal Prediction

We introduce a distributed spatio-temporal artificial neural network arc...
research
09/21/2020

Hidden Latent State Inference in a Spatio-Temporal Generative Model

Knowledge of the hidden factors that determine particular system dynamic...
research
12/23/2019

Multi-level Convolutional Autoencoder Networks for Parametric Prediction of Spatio-temporal Dynamics

A data-driven framework is proposed for the predictive modeling of compl...
research
01/04/2021

Conditional Local Filters with Explainers for Spatio-Temporal Forecasting

Spatio-temporal prediction is challenging attributing to the high nonlin...
research
10/29/2021

Deep inference of latent dynamics with spatio-temporal super-resolution using selective backpropagation through time

Modern neural interfaces allow access to the activity of up to a million...
research
01/17/2022

Learning Wave Propagation with Attention-Based Convolutional Recurrent Autoencoder Net

In this paper, we present an end-to-end attention-based convolutional re...

Please sign up or login with your details

Forgot password? Click here to reset