We see right into the heart of how random bipartite affilation networks work, laying out the essential calculations as natural generalizations of their random network counterparts. Essentially, we hop back and forth between two kinds of degree distributions. We then explore some earlier observations of how well random affiliations perform in modeling real bipartite networked systems (pretty good). We conclude with the most basic kind of spreading: diffusion. We let a random walker loose on a network and see that the long term location probability matches up with standard diffusion. The walker will be at node with probability proportional to the node's degree, leading to the important observation that uniformity of stuff is in the transporation along edges. It's always about the edges.

Prediction and basketball games with highly unlikely finishes; the tarot cards take off (a little bit).

Back of the envelope notes made during class: Pages 1 to 4 and Pages 5 and 6.

2016/03/22

1:19:43