Bubble Entropy
An Entropy almost free of Parameters
A critical point in
any definition of entropy is the selection of the parameters employed to obtain
an estimate in practice. We propose a new definition of entropy aiming to
reduce the significance of this selection. We call the new definition Bubble
Entropy.
In this definition, we
embed the time series in the m-dimensional space. We use the bubble sort
algorithm to sort each vector in the m-dimensional space and count the number
of swaps performed for each vector. Doing so, we create a more coarse-grained
distribution and then compute the entropy of this distribution.
The definition
proposed is almost free of parameters. The most common ones are the scale
factor r and the embedding dimension m. In our definition, the scale factor
is totally eliminated and the importance of m
is significantly reduced. The proposed method presents increased stability and discriminating
power.
Contact: manis -at-
cs -dot- uoi -dot- gr
Selected Publications:
-
George Manis, Md Aktaruzzaman and Roberto Sassi, “Bubble entropy: an entropy almost free of
parameters,” IEEE Transactions on Biomedical Engineering,
vol. 64, no. 11, pp. 1558–2531, Nov. 2017 [link]
-
George Manis and Roberto Sassi, “Tolerance to spikes: a
comparison of sample and bubble entropy,’’ in Proc. of
Computing in Cardiology, Rennes, France, 2017 [link]