Fast Computation of Entropy
Reducing computation time for entropy estimation in m-dimensional space
Most popular
definitions of entropy are computationally expensive methods which may require a
large amount of time when used in long series or with a large number of
signals. The computationally intensive part of the computation is the
similarity check between points in the m-dimensional
space. We propose algorithms aiming to compute entropy fast. All algorithms
return exactly the values of the metric definition, no approximation techniques
are used.
The key idea is based
on avoiding unnecessary comparisons by detecting them early. We use the term unnecessary to refer to those
comparisons for which we know a priori that they will fail at the similarity
check. The number of avoided comparisons is proved to be very large, resulting
in an analogous large reduction of execution time.
Contact: manis -at- cs -dot- uoi -dot- gr
Selected Publications:
-
George Manis, Md. Aktaruzzaman,
and Roberto Sassi, “Low computational cost for
sample entropy,” Entropy, MDPI, vol. 20, art. 61, 2018 [link]
-
George Manis, “Fast computation of
approximate entropy,” Computer Methods and Programs in
Biomedicine, Elsevier, vol. 91,
no. 1, pp. 48–54, Jul. 2008 [link]