Code and data from: Gaussian Process Regression for Estimating EM Ducting Within the Marine Atmospheric Boundary Layer

dc.contributor.authorSit, Hilarie
dc.contributor.authorEarls, Christopher J
dc.descriptionRedistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. * Neither the name of the copyright holders nor the name of Cornell University may be used to endorse or promote products derived from this software without specific prior written permission. Private, research, and institutional usage is without charge. Distribution of modified versions of this soure code is admissible UNDER THE CONDITION THAT THIS SOURCE CODE REMAINS UNDER COPYRIGHT OF THE ORIGINAL DEVELOPERS, BOTH SOURCE AND OBJECT CODE ARE MADE FREELY AVAILABLE WITHOUT CHARGE, AND CLEAR NOTICE IS GIVEN OF THE MODIFICATIONS. Distribution of this code as part of a commercial system is permissible ONLY BY DIRECT ARRANGEMENT WITH THE DEVELOPERS. THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.en_US
dc.description.abstractWe show that Gaussian process regression (GPR) can be used to infer the electromagnetic (EM) duct height within the marine atmospheric boundary layer (MABL) from sparsely sampled propagation factors within the context of bistaticradars. These propagation factors are simulated using PETOOL, developed by Ozgun et al. 2011, and the datasets for the three cases that correspond to the different sparse sampling techniques can be found in the data folder. We use GPR to calculate the posterior predictive distribution on the labels (i.e. duct height) from both noise-free and noise-contaminated array of propagation factors. For duct height inference from noise-contaminated propagation factors, we compare a naive approach, utilizing one random sample from the input distribution (i.e. disregarding the input noise), with an inverse-variance weighted approach, utilizing a few random samples to estimate the true predictive distribution. The resulting posterior predictive distributions from these two approaches are compared to a "ground truth" distribution, which is approximated using a large number of Monte-Carlo samples. We use Python 3.6.4 and scikit-learn 0.20.2. The ability of GPR to yield accurate duct height predictions using few training examples, along with its inference speed, indicates the suitability of the proposed method for real-time applications. This is the dataset and code that supports this work.en_US
dc.description.sponsorshipThe authors gratefully acknowledge ONR Division 331 and Dr. Steve Russell for the financial support of this work through grant N00014-19-1-2095.en_US
dc.relation.isreferencedbySit, Hilarie, and Christopher J. Earls. “Gaussian Process Regression for Estimating EM Ducting Within the Marine Atmospheric Boundary Layer.” Radio Science, vol. 55, no. 6, June 2020.
dc.subjectgaussian process regressionen_US
dc.subjectmarine atmospheric boundary layeren_US
dc.subjectevaporation ductsen_US
dc.subjectinverse problemen_US
dc.subjectmachine learningen_US
dc.titleCode and data from: Gaussian Process Regression for Estimating EM Ducting Within the Marine Atmospheric Boundary Layeren_US


Original bundle
Now showing 1 - 2 of 2
No Thumbnail Available
4.44 KB
Unknown data format
No Thumbnail Available
310.71 MB
Data Compression Utility