Affinity propagation for clustering

Basic idea: message passing. Number of clusters automatically determined.

2 matrices to be updated: responsibility R and availability A.

R: to let exemplars compete for ownership of a data point.

A; to gather evidence from data points about whether each candidate exemplar would be a good one.

r(i,k) reflects the accumulated evidence for how well-suited point k is to serve as the exemplar for point i, taking into account other potential exemplars for point i.

a(i,k) reflects the accumulated evidence for how appropriate it would be for point i to choose point k as its exemplar, taking into account the support from other points that point k should be an exemplar.

For point i, the value of k that maximize a(i,k)+r(i,k) either identifies point i as an exemplar if k=i, or identifies the data point that is the exemplar for point i.