New method estimates Fourier transforms from finite data without periodicity assumptions.
problem Estimating Fourier transforms from discrete data points without periodicity assumptions.
method Gaussian process regression with gradient ascent method to estimate covariance function.
result Sharp and precise estimation of spectral density in noise-free and noisy signals.