Abstract
Transfer entropy is a measure of causality that has been widely applied and one of its identities is the sum of mutual information terms. In this article we evaluate two existing methods of mutual information estimation in the specific application of detecting causality between a discrete random process and a continuous random process: binning method and nearest neighbours method. Simulated examples confirm, in the overall scenario, that the nearest neighbours method detects causality more reliably than the binning method.
Highlights
Transfer entropy (TE), as well as Granger causality and directed information, is a measure of causality
The contribution of this paper is to investigate the application of both binning method and nearest neighbours method described above for mutual information, in equation
In order to evaluate the performance of these methods, we have developed some examples involving causality in mixed cases
Summary
Transfer entropy (TE), as well as Granger causality and directed information, is a measure of causality. The purpose of this paper is to evaluate TE estimators for these mixed cases, which may be of interest for those working with mixed processes and with a causality measure necessity. This may be relevant in the context of a Poisson channel with feedback. Two methods of estimation are explored in this paper for this case of mixed processes. Both methods stems from an identity for TE, written as a sum of two mutual information terms.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
