Entropy (Sep 2022)

Decoding ‘Maximum Entropy’ Deconvolution

  • Long V. Le,
  • Tae Jung Kim,
  • Young Dong Kim,
  • David E. Aspnes

DOI
https://doi.org/10.3390/e24091238
Journal volume & issue
Vol. 24, no. 9
p. 1238

Abstract

Read online

For over five decades, the mathematical procedure termed “maximum entropy” (M-E) has been used to deconvolve structure in spectra, optical and otherwise, although quantitative measures of performance remain unknown. Here, we examine this procedure analytically for the lowest two orders for a Lorentzian feature, obtaining expressions for the amount of sharpening and identifying how spurious structures appear. Illustrative examples are provided. These results enhance the utility of this widely used deconvolution approach to spectral analysis.

Keywords