In our previous paper, we investigate the consensus problem with von Mises distribution. We want to extend our analysis to more general distributions, in particular the von Mises-Fisher (vMF) distribution over hyperspheres.

In fact, vMF distributions are used in some machine learning algorithms. In common machine learning applications, algorithms handle feature vectors. When the feature vector is normalized as a unit vector, vMF distributions are usually a very good option to model such distribution. Therefore, researchers that use vMF distributions say that vMF distributions are important for feature vectors with directional information but not norm information.

I have seen applications in text analysis, face recognition, gene expression data. The following figure shows the flow chart of a face recognition process.

[Hasnat et al., 2017]

However, most of the machine learning works are in clustering and learning the vMF distribution. As our contributions will be in the fusion and network consensus scenario. Maybe there is an interesting fusion scheme that we can use our algorithm.

Next Post Previous Post