An increasing number of studies across many research fields from biomedical engineering to finance are employing measures of entropy to quantify the regularity, variability or randomness of time series and image data.
How to measure entropy? How to find an accurate tool that is accessible to the greatest?
Matthew Flood, a postdoctoral fellow at the Luxembourg Institute of Health, has developed an open source toolkit for entropic time series analysis dedicated to students, researchers and companies. Its name is EntropyHub.
Entropy as a measure of the disorder of the universe
The world around us is composed of systems that exhibit nonlinear dynamical behaviour, e.g. the human body, the financial markets, and the climate.
Measuring the uncertainty of their behaviour is crucial for predicting events where these systems malfunction. Entropy allows us to measure that uncertainty.
In recent years, entropy has emerged as a complexity measure for the study of time series from biological systems, including the brain, heart, or muscles. Entropy measures have been used to analyse short, sparse and noisy medical time series.
The concept of entropy is also used in many other fields of science like economics or sociology.
EntropyHub: a new hope
While entropy is gaining momentum, there is a lack of validated, open source software tools that allow researchers to apply these methods. To date, software packages for performing entropy analysis are often run using graphical interfaces, lack the necessary documentation, or do not include functions for more advanced entropy methods.
This is what Matthew Flood experienced during his doctoral studies while looking for resources. He started to write algorithms to have a set of tools for himself in his daily job. Compiling toolboxes, Matthew Flood was able to compare different entropic series.
“Despite the popularity of entropy analysis in research, there was nowhere online to find comprehensive set of validated algorithms for calculating different entropy measures. Many researchers and students contacted me to ask where they can find a particular piece of code or function. Indeed, it’s not the easiest thing to write since you have to be absolutely certain it is accurate”Matthew Flood
As a postdoctoral fellow in Research Luxembourg, Matthew received the time and space to design an online solution. Dubbed EntropyHub, this open-source toolkit is designed to perform entropic time series analysis in MATLAB, Python and Julia, i.e. numerical programming languages.
EntropyHub provides an extensive range of more than forty functions for estimating cross-, multiscale, multiscale cross-, and bidimensional entropy, each including a number of keyword arguments that allows the user to specify multiple parameters in the entropy calculation.
An example of inclusive science
EntropyHub provides tools that make advanced entropic time series analysis straightforward and reproducible. Matthew Flood believes in inclusive science. Hence, EntropyHub is accessible and results are reproductible.
“I’m an advocate of well designed and well developed open source software tools. If you are able to share the tools you use with the community, it also means that the results of your research can be replicated. If you are more open about your work, it becomes more reliable. This is what we are aiming for.”Matthew Flood
The software is already gaining traction as Matthew is in contact with researchers in Brazil, Columbia, Ireland and Portugal.
How can technological solutions provide clinicians and patients with effective treatment outcomes for orthopaedic injuries and diseases? What role can motion capture technology play in delivering quantitative clinical diagnoses of sport-related injuries? How can researchers harness the data from wearable sensors and digital devices to understand the impact of impaired mobility during activities of daily…Keep reading