Coupling physics based simulations with Artificial Intelligence
Enhancing the current state of the art simulations for Space Weather with prior knowledge gathered from historical satellite data, through a portfolio of data enhanced reduced models.
A data driven approach to Space Weather
In this project we aim to enhance the current state of the art simulations for Space Weather, by using prior knowledge gathered from historical satellite data. Several Machine Learning techniques are used for data mining, classification and regression. The long term objective of the project is the creation of a portfolio of data enhanced reduced models, along with automated rules for model selection. Depending on the real time conditions observed by satellites, the resulting grey box model should choose the relative importance between physical and empirical estimations. Coupling physics based simulations with Artificial Intelligence can provide powerful insights into patterns in data and support predictive models for forecasting behavior and preferences.
Data mining
Prior knowledge drawn from historical satellite data guides every estimate.
Classification
Machine learning methods separate observed conditions into meaningful regimes.
Regression
Predictive models estimate the quantities that physics alone cannot capture.
Grey box model
A model that weighs physical and empirical estimates against real time conditions.
CWI/INRIA consortium
This project started as funded by a CWI/INRIA collaboration. Several other parties have since joined in this activity, either as external collaborators or as more actively involved partners. Visit the team page for a list of all the people involved.
CWI
CWI is the Dutch National Center for Mathematics and Computer Science.
INRIA
INRIA is the French Institute for Research in Computer Science and Automation.
Read the emerging cyberthreats whitepaper
A companion publication on how privileged access management helps organisations stay ahead of continuously shifting cybersecurity threats.