Signal Estimation of Time Series by Fourier Analysis 


Fourier analysis is a method used for signal estimation, which separates an underlying signal 
from noise, of discrete time series data with cyclical patterns. The methodology behind this 
analysis will be explained, focusing on finding multiple periods.  Analysis will be done on the 
subject level as well as the group level.  Once the methods are described, a simulation will be 
done for use as a comparison to two other examples.  Using these two biomedical time series 
examples, one on gestational diabetes and the other on Herpes Simplex Virus, signal estimation 
will be performed and compared to the results of the simulation.