Search academic texts

Information × Registration Number 0222U000457, 0121U112420 , R & D reports Title Investigation of machine learning applicability for detection of traveling ionospheric disturbances popup.stage_title Head Bohomaz Oleksandr V., Кандидат технічних наук Registration Date 10-01-2022 Organization Institute of Ionosphere popup.description2 The data of HF sounding and incoherent scattering, obtained in Antarctica and Ukraine during several years for different seasons, for which various known algorithms of machine learning were applied, analyzed and chosen. It has been shown that the most effective method for detecting TIDs is deep learning based on the use of artificial neural networks consisting of more than three layers. A technique of HF sounding data processing for further use of machine learning algorithms has been developed. It is proved that for effective identification of TIDs it is necessary to conduct a joint analysis of altitude-time variations of relative changes in IS signal power (electron density) and the results of spectral analysis at a number of altitudes. Product Description popup.authors Bogomaz Oleksandr V. Zhivolup Taras G. Kotov Dmytro V. Panasenko Serhii V Reznychenko Maryna O. popup.nrat_date 2022-03-09 Close