scholarly journals Unsupervised Solar Wind Classification Using Wavelet Variational Autoencoders and Self-Organazing Maps

2021 ◽  
Author(s):  
Jorge Amaya ◽  
Sara Jamal ◽  
Giovanni Lapenta
Keyword(s):  
Solar Wind ◽  
Single Point ◽  
Coronal Holes ◽  
Principal Component ◽  
Data Transformation ◽  
Unsupervised Classification ◽  
Kernel Principal Component Analysis ◽  
Horizon 2020 ◽  
Research And Innovation

<p>Last year we published an automatic method for the automatic classification of the solar wind [1]. We showed that data transformation and unsupervised clustering can be used to classify observations made by the ACE spacecraft. Two data transformation techniques were used: Kernel Principal Component Analysis (KPCA) and Auto-encoder Neural Networks. After data transformation three clustering techniques were tested: k-means, Bayesian Gaussian Mixtures (BGM), and Self-Organizing Maps (SOM). Although the results were very positive we ran into a few difficulties: a) the data from the ACE mission contains a very small population of observations originated at high latitude coronal holes, b) the measured features contain a high degree of intercorrelation, c) the data distribution is compact in the feature space, and d) the final algorithm produces a single categorical class for a single point in time.</p><p><br>In this work we present an improvement of the model that redresses some of the limitations above. We are still making use of the two main features of our previous work, i.e. the data transformation using auto-encoders and the unsupervised classification using SOM. But in the present work: a) we include the analysis of Ulysses data with observations of the solar wind originated at high latitudes; b) we perform a Factor Analysis to reduce the number of features used as inputs; c) we transform windows of time of the multi-variate time series (instead of instantaneous observations) into scalograms using wavelet transformations; d) we apply the variational version of the auto-encoder [2] to parametrize the scalograms; f) we finally use the SOM to automatically classify the windows of time in different categories.</p><p><br>This method can be adapted to the classification of observations from the Parker Solar Probe and Solar Orbiter missions.</p><p><br>The work presented in this abstract has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No. 754304 (DEEP-EST, www.deep-projects.eu), and from the European Union's Horizon 2020 research and innovation programme under grant agreement No 776262 (AIDA, www.aida-space.eu).</p><p><br>[1] Amaya, Jorge, Romain Dupuis, Maria Elena Innocenti, and Giovanni Lapenta. "Visualizing and Interpreting Unsupervised Solar Wind Classifications." arXiv preprint arXiv:2004.13430 (2020).</p><p>[2] Kingma, Diederik P., and Max Welling. "Auto-encoding variational bayes." arXiv preprint arXiv:1312.6114 (2013).</p>

Sun Planet Interactions Digital Environment on Request (SPIDER) for Europlanet RI H2024

2020 ◽  
Author(s):  
Nicolas André ◽  
Vincent Génot ◽  
Andrea Opitz ◽  
Baptiste Cecconi ◽  
Nick Achilleos ◽  
...  
Keyword(s):  
Solar Wind ◽  
Giant Planet ◽  
High Energy ◽  
Research Infrastructure ◽  
Radiation Environment ◽  
High Energy Particle ◽  
Particle Tracing ◽  
Horizon 2020 ◽  
Research And Innovation ◽  
Infrastructure Project

<p>The H2020 Europlanet-2020 programme, which ended on Aug 31<sup>st</sup>, 2019, included an activity called PSWS (Planetary Space Weather Services), which provided 12 services distributed over four different domains (A. Prediction, B. Detection, C. Modelling, D. Alerts) and accessed through the PSWS portal (http://planetaryspaceweather-europlanet.irap.omp.eu/):</p> <p>A1. 1D MHD Solar Wind Prediction Tool – HELIOPROPA,</p> <p>A2. Propagation Tool,</p> <p>A3. Meteor showers,</p> <p>A4. Cometary tail crossings – TAILCATCHER,</p> <p>B1. Lunar impacts – ALFIE,</p> <p>B2. Giant planet fireballs – DeTeCt3.1,</p> <p>B3. Cometary tails – WINDSOCKS,</p> <p>C1. Earth, Mars, Venus, Jupiter coupling- TRANSPLANET,</p> <p>C2. Mars radiation environment – RADMAREE,</p> <p>C3. Giant planet magnetodiscs – MAGNETODISC,</p> <p>C4. Jupiter’s thermosphere, D. Alerts.</p> <p>In the framework of the starting Europlanet-2024 programme, SPIDER will extend PSWS domains (A. Prediction, C. Modelling, E. Databases) services and give the European planetary scientists, space agencies and industries access to 6 unique, publicly available and sophisticated services in order to model planetary environments and solar wind interactions through the deployment of a dedicated run on request infrastructure and associated databases.</p> <p>C5. A service for runs on request of models of Jupiter’s moon exospheres as well as the exosphere of Mercury,</p> <p>C6. A service to connect the open-source Spacecraft-Plasma Interaction Software (SPIS) software with models of space environments in order to compute the effect of spacecraft potential on scientific instruments onboard space missions. Pre-configured simulations will be made for Bepi-Colombo and JUICE missions,</p> <p>C7. A service for runs on request of particle tracing models in planetary magnetospheres,</p> <p>E1. A database of the high-energy particle flux proxy at Mars, Venus and comet 67P using background counts observed in the data obtained by the plasma instruments onboard Mars Express (operational from 2003), Venus Express (2006–2014), and Rosetta (2014–2015);</p> <p>E2. A simulation database for Mercury and Jupiter’s moons magnetospheres and link them with prediction of the solar wind parameters from Europlanet-RI H2020 PSWS services.</p> <p>A1. An extension of the Europlanet-RI H2020 PSWS Heliopropa service in order to ingest new observations from Solar missions like the ESA Solar Orbiter or NASA Solar Parker Probe missions and use them as input parameters for solar wind prediction;</p> <p>These developments will be discussed in the presentation.</p> <p>The Europlanet 2020 Research Infrastructure project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 654208.</p> <p>The Europlanet 2024 Research Infrastructure project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 871149.</p>

scholarly journals Pre-mRNA processing entropy in a mouse model of trauma

2021 ◽  
Author(s):  
Maximilian S. Jentzsch ◽  
Alger M. Fredericks ◽  
Jason T. Machan ◽  
Alfred Ayala ◽  
Sean F. Monaghan
Keyword(s):  
Critical Illness ◽  
Cecal Ligation ◽  
Principal Component ◽  
Unsupervised Classification ◽  
Mrna Processing ◽  
Small Set ◽  
Processing Event ◽  
Exposure To Trauma ◽  
Generation Sequencing

AbstractPurposeNext generation sequencing has expanded our understanding of many disease processes, including trauma and critical illness. Many studies focus identifying a small set of genes or proteins that are aberrantly expressed. Our objective was to determine whether global differences in pre-mRNA processing entropy, or disorder, could offer novel insights in the setting of critical illness.MethodsWe used an established murine model of trauma that consisted of hemorrhagic shock and cecal ligation and puncture. In our first experiment mice exposed to trauma were compared to controls. In our second experiment, survival 14 days after exposure to trauma was studied. Using deep RNA sequencing we determined entropy values for every pre-mRNA processing event identified. We then used principal component analysis (PCA) to conduct unsupervised classification of the data.ResultsMice exposed to trauma separated from controls using PCA. Similarly, mice that did not survive 14 days post exposure clustered closely together on PCA.ConclusionOur results suggest that there is a substantial difference in global pre-mRNA processing entropy in mice exposed to trauma vs. controls, and that pre-mRNA processing entropy may be helpful in predicting mortality. The method introduced here is easily transferrable to other disease processes and samples.

scholarly journals Sequential PCA-based Classification of Mediterranean Forest Plants using Airborne Hyperspectral Remote Sensing

2019 ◽  
Vol 11 (23) ◽  
pp. 2800
Author(s):  
Alon Dadon ◽  
Moshe Mandelmilch ◽  
Eyal Ben-Dor ◽  
Efrat Sheffer
Keyword(s):  
Remote Sensing ◽  
Principal Component ◽  
Unsupervised Classification ◽  
Hyperspectral Remote Sensing ◽  
Mediterranean Forest ◽  
Full Potential ◽  
Sequential Image ◽  
Forest Plants ◽  
Intensive Sampling

In recent years, hyperspectral remote sensing (HRS) has become common practice for remote analyses of the physiognomy and composition of forests. Supervised classification is often used for this purpose, but demands intensive sampling and analyses, whereas unsupervised classification often requires information retrieval out of the large HRS datasets, thereby not realizing the full potential of the technology. An improved principal component analysis-based classification (PCABC) scheme is presented and intended to provide accurate and sequential image-based unsupervised classification of Mediterranean forest species. In this study, unsupervised classification and reduction of data size are performed simultaneously by applying binary sequential thresholding to principal components, each time on a spatially reduced subscene that includes the entire spectral range. The methodology was tested on HRS data acquired by the airborne AisaFENIX HRS sensor over a Mediterranean forest in Mount Horshan, Israel. A comprehensive field-validation survey was performed, sampling 257 randomly selected individual plants. The PCABC provided highly improved results compared to the traditional unsupervised classification methodologies, reaching an overall accuracy of 91%. The presented approach may contribute to improved monitoring, management, and conservation of Mediterranean and similar forests.

Improving CME evolution and arrival predictions with AMR and grid stretching in EUHFORIA

2021 ◽  
Author(s):  
Tinatin Baratashvili ◽  
Christine Verbeke ◽  
Nicolas Wijsen ◽  
Emmanuel Chané ◽  
Stefaan Poedts
Keyword(s):  
Solar Wind ◽  
Adaptive Mesh Refinement ◽  
Adaptive Mesh ◽  
Mhd Equations ◽  
Efficiency Gain ◽  
The European Union ◽  
Interplanetary Shocks ◽  
Horizon 2020 ◽  
Research And Innovation ◽  
Speed Up

<p>Coronal Mass Ejections (CMEs) are the main drivers of interplanetary shocks and space weather disturbances. Strong CMEs directed towards Earth can cause severe damage to our planet. Predicting the arrival time and impact of such CMEs can enable to mitigate the damage on various technological systems on Earth. </p><p>We model the inner heliospheric solar wind and the CME propagation and evolution within a new heliospheric model based on the MPI-AMRVAC code. It is crucial for such a numerical tool to be highly optimized and efficient, in order to produce timely forecasts. Our model solves the ideal MHD equations to obtain a steady state solar wind configuration in a reference frame corotating with the Sun. In addition, CMEs can be modelled by injecting a cone CME from the inner boundary (0.1 AU).</p><p>Advanced techniques, such as grid stretching and Adaptive Mesh Refinement (AMR) are employed in the simulation. Such methods allow for high(er) spatial resolution in the numerical domain, but only where necessary or wanted. As a result, we can obtain a detailed, highly resolved image at the (propagating) shock areas, without refining the whole domain.</p><p>These techniques guarantee more efficient simulations, resulting in optimised computer memory usage and a significant speed-up. The obtained speed-up, compared to the original approach with a high-resolution grid everywhere, varies between a factor of 45 - 100 depending on the domain configuration. Such efficiency gain is momentous for the mitigation of the possible damage and allows for multiple simulations with different input parameters configurations to account for the uncertainties in the measurements to determine them. The goal of the project is to reproduce the observed results, therefore, the observable variables, such as speed, density, etc., are compared to the same type of results produced by the existing (non-stretched, single grid) EUropean Heliospheric FORecasting Information Asset (EUHFORIA) model and observational data for a particular event on 12th of July, 2012. The shock features are analyzed and the results produced with the new heliospheric model are in agreement with the existing model and observations, but with a significantly better performance. </p><p> </p><p><strong>This research has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 870405 (EUHFORIA 2.0).</strong></p>

Modeling the Sun – Earth propagation of solar disturbances for the H2020 SafeSpace project

2021 ◽  
Author(s):  
Benoit Lavraud ◽  
Rui Pinto ◽  
Rungployphan Kieokaew ◽  
Evangelia Samara ◽  
Stefaan Poedts ◽  
...  
Keyword(s):  
Solar Wind ◽  
Interplanetary Space ◽  
Coronal Magnetic Field ◽  
Physical Parameters ◽  
The European Union ◽  
Horizon 2020 ◽  
Research And Innovation ◽  
Reconstruction Methods ◽  
Solar Disturbances ◽  
Magnetic Field Computation

<p>We present the solar wind forecast pipeline that is being implemented as part of the H2020 SafeSpace project. The Goal of this project is to use several tools in a modular fashion to address the physics of Sun – interplanetary space – Earth’s magnetosphere. This presentation focuses on the part of the pipeline that is dedicated to the forecasting – from solar measurements – of the solar wind properties at the Lagrangian L1 point. The modeling pipeline puts together different mature research models: determination of the background coronal magnetic field, computation of solar wind acceleration profiles (1 to 90 solar radii), propagation across the heliosphere (for regular solar wind, CIRs and CMEs), and comparison to spacecraft measurements. Different magnetogram sources (WSO, SOLIS, GONG, ADAPT) can be combined, as well as coronal field reconstruction methods (PFSS, NLFFF), wind (MULTI-VP) and heliospheric propagation models (CDPP 1D MHD, EUHFORIA). We aim at providing a web-based service that continuously supplies a full set of bulk physical parameters of the solar wind at 1 AU several days in advance, at a time cadence compatible with space weather applications. This work has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 870437.</p>

Laser-Welding Spots Detection Based on Original Moment Values

2014 ◽  
Vol 599-601 ◽  
pp. 974-980
Author(s):  
Xiao Long Qi ◽  
Bin Fang ◽  
Shu Mei Wang
Keyword(s):  
Principal Component Analysis ◽  
Laser Welding ◽  
Image Classification ◽  
Principal Component ◽  
Component Analysis ◽  
Kernel Principal Component Analysis ◽  
Invariant Moments ◽  
The Past

In the past decades, the theories of invariant moments have been researched extensively and wildly used in many fields. However, for the laser-welding spots of titanium tubes or other fixed objects, the invariant moments are inapplicable. Besides, the studies and experiments about image classification by means of the original moment values were barely proposed. In this paper, the method of classification based on original moment values is introduced, and an improved approach of KPCA (kernel principal component analysis) in order to reduce the inner-class distance of the qualified laser-welding spots is also discussed. Finally, experiments are carried out to validate the classification ability, and results show that the original moment values are suited as pattern features in classification of fixed objects.

Sign in / Sign up

Export Citation Format

Share Document