Reply to "Comment on `Performance of different synchronization measures in real data: A case study on electroencephalographic signals'"

Abstract

We agree with the Comment by Nicolaou and Nasuto about the utility of mutual information (MI) when properly estimated and we also concur with their view that the estimation based on k nearest neighbors gives optimal results. However, we claim that embedding parameters can indeed change MI results, as we show for the electroencephalogram data sets of our original study and for coupled chaotic systems. Furthermore, we show that proper embedding can actually improve the estimation of MI with the k nearest neighbors algorithm.