Modeling the Location and Shape of the Magnetopause Using Machine Learning Methods

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F23%3A10477991" target="_blank" >RIV/00216208:11320/23:10477991 - isvavai.cz</a>
Result on the web
<a href="https://physics.mff.cuni.cz/wds/proc/pdf23/WDS23_09_f2_AghabozorgiNafchi.pdf" target="_blank" >https://physics.mff.cuni.cz/wds/proc/pdf23/WDS23_09_f2_AghabozorgiNafchi.pdf</a>
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
Modeling the Location and Shape of the Magnetopause Using Machine Learning Methods
Original language description
Empirical models for predicting the location of the magnetopause currentlyin use typically rely on the identification of individual magnetopause crossings, and theirfitting by a predefined magnetopause shape. Although the assumed analytical shape formulamay be rather complicated and general, it represents a principal apriori limitation of themodel. We remove this limitation by applying an approach based on an artificial neuralnetwork. A large data set of about 15,000 subsolar magnetopause crossings is used for thetraining, resulting in a direct data-driven model predicting the magnetopause radial distanceas a function of relevant solar wind parameters without any additional assumptions. Themodel performance is evaluated using the testing data set of magnetopause crossings andby a comparison with a former widely used empirical model.
Czech name
—
Czech description
—

Project
—
Continuities
S - Specificky vyzkum na vysokych skolach<br>I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Publication year
2023
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection
WDS'23 Proceedings of Contributed Papers - Physics
ISBN
978-80-7378-503-1
ISSN
—
e-ISSN
—
Number of pages
7
Pages from-to
71-77
Publisher name
Matfyzpress
Place of publication
Praha
Event location
Praha
Event date
May 30, 2023
Type of event by nationality
EUR - Evropská akce
UT code for WoS article
—

Similar results(10)