The Implementation of Neural Networks for Phaseless Parametric Inversion
| Autor: | Joe LoVetri, Ian Jeffrey, Colin Gilmore, Vahab Khoshdel, Ryan Kruk, Keeley Edwards, Kennedy Krakalovich |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Artificial neural network
Computer science 0211 other engineering and technologies Inverse transform sampling Experimental data Inversion (meteorology) 02 engineering and technology Synthetic data Bin 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Ligand cone angle Algorithm 021101 geological & geomatics engineering Parametric statistics |
| Zdroj: | 2020 XXXIIIrd General Assembly and Scientific Symposium of the International Union of Radio Science. |
| DOI: | 10.23919/ursigass49373.2020.9232216 |
| Popis: | We present a machine learning work flow for the parametric inversion of grain bin measurements in which a neural network is trained solely on synthetic data for a unique bin geometry. This neural network can subsequently be used to rapidly obtain 4 inversion parameters (grain height, cone angle, and bulk real and imaginary permittivity of the grain) from uncalibrated, experimental data. We have previously shown that these 4 parameters can be used to calibrate experimental data and serve as prior information for full-data inversion. Our results show that a densely connected neural network that supports multifrequency data can better predict the cone angle of grain, and perform almost as well on grain height predictions, as the single-frequency simplex inversion method previously described. These findings suggest that neural networks trained on synthetic data may be a useful tool in the inversion of experimental data, providing prior information and a method for calibration. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|