2014 Volume 13 Issue 3 Pages 190-192
Maximum entropy method (MEM) is one of the key techniques for spectral analysis. The main feature is improving spectral resolution with short time-series data. We adopt MEM to the calculation of time-dependent density functional theory (TDDFT) in real time, which is intensively studied and applied to computing optical properties. Usually, the spectrum is obtained from Fourier transform (FT) of the time evolution of the dipole moment. In the spectrum comparison of MEM and FT calculations for several molecules, we realize that MEM provides a similar spectrum to that of FT with almost half of the time evolution. In the MEM analysis, not only the total number of time-series data, but also the length of the autocorrelation has an important role to obtain the high resolution spectrum. In this paper, we present the effectiveness and characteristic features of MEM, comparing with those of the FT technique in TDDFT.
