There’s been some new progress on graph Laplacians!
As a mathematical physicist, I’ve always been in love with the Laplacian:
$latex \displaystyle{ \nabla^2 = \frac{\partial^2}{\partial x^2} + \frac{\partial^2}{\partial y^2} + \frac{\partial^2}{\partial z^2} }$ 
It shows up in many of the most fundamental equations of physics: the wave equation, the heat equation, Schrödinger’s equation… and Poisson’s equation:
$latex \nabla^2 \phi = \rho $
which says how the density of matter, $latex \rho,$ affects the gravitational potential $latex \phi.$
As I’ve grown interested in network theory, I’ve gotten more and more interested in ‘graph Laplacians’. These are discretized versions of the Laplacian, where we replace 3-dimensional space by a ‘graph’, meaning something like this:
You can get a lot of interesting information about a graph from its Laplacian. You can also set up discretized versions of all the famous equations I mentioned.
The new progress is a
View original post 540 more words
