Temporal Data Mining

206    ◾    Temporal Data Mining

histogram as well as the skewness and kurtosis of the histograms of the

wavelet coecients. e Bayesian was successful in classifying the data

sets of the testing set in the correct class.

In [Mon02], the authors propose a method for instantaneous parameter

estimation in cardiovascular time series by harmonic and time frequency

analysis. ere are two well-known methods, smoothed pseudo Wigner–

Ville distribution (SPWVD) and complex demodulation (CDM), to be used

as time-varying spectral estimators of physiological signals. ese meth-

ods provide very similar results in real cardiovascular time series, but they

also have complementary properties:

Phase relationship between respiratory activity and high frequency ...