LAUSR

A multiple regression and machine learning approach is proposed to solve multiple plasma parameter inversion in complex systems. For an electromagnetic (EM) wave and plasma interaction system, it is common to retrieve plasma parameters based on the reflected or scattered EM field, whereas multiple parameter inversion becomes particularly challenging for anisotropic inhomogeneous plasma medium. Artificial neural network (ANN) is applied to retrieve the plasma density fluctuation in inhomogeneous medium. The method is computationally efficient and robust based solely on time series of data collected at a fixed number of spatial locations. The inversion algorithm opens venues and application for inhomogeneous plasma parameter inversion for EM wave and plasma interaction system at different frequencies.