Reconstruction of turbulence time series in a statistically stationary state is discussed by using a machine learning algorithm. We use data obtained by Langmuir probes in the Plasma Assembly for Nonlinear Turbulence Analysis (PANTA). It is shown that even if the distance between two probes is not adequate to resolve the turbulence, the nonlinear regression via the machine learning can give reconstruction better than those by the linear regression and the linear interpolation. Wave forms and frequency spectra show that drift waves are well reconstructed by the machine learning.
All Science Journal Classification (ASJC) codes
- Condensed Matter Physics