News for hackers. Stuff that matters.
Re:The Unreasonable Effectiveness of Spectral Graph Theory: ... by visarga on Sunday November 15, 2015 @03:12PM (#10570961)
What are the practical uses for this? I thought it was going to be about PageRank.
Re:The Unreasonable Effectiveness of Spectral Graph Theory: ... by fleitz on Sunday November 15, 2015 @04:51AM (#10569109)
Is it just me or would this not drastically speed up calculation of the weights for neural nets?
Re:The Unreasonable Effectiveness of Spectral Graph Theory: ... by hiddencost on Sunday November 15, 2015 @06:04AM (#10569271)
SVD is in fact used to compress dnns.

Of course, like many things in ML, as soon as you make it faster, you just make the model bigger and turn the speed gains into accuracy gains.

Re:The Unreasonable Effectiveness of Spectral Graph Theory: ... by zzleeper on Sunday November 15, 2015 @04:56AM (#10569124)
Well the Ax=b part is quite general as long as A has a special structure (diagonally dominant).

Good luck coding it though, I'm still looking for an implementation of a low stretch spanning tree (needed in order for the solver to be really lineaR)

Re:The Unreasonable Effectiveness of Spectral Graph Theory: ... by espeed on Monday November 16, 2015 @12:11PM (#10575584)
"Using Petal-Decompositions to Build a Low Stretch Spanning Tree" (http://research.microsoft.com/pubs/163400/spanning-full1.pdf):

  Stretch: O(log n log log n)
  Time:    O(m log n log log n)
Re:The Unreasonable Effectiveness of Spectral Graph Theory: ... by mike_mg on Sunday November 15, 2015 @10:50AM (#10569904)
Could you please explain me how? I watched the lecture, but can't make connection. I understand how neural networks work and how they are optimized. I'm stupid. Thanks
Re:The Unreasonable Effectiveness of Spectral Graph Theory: ... by fleitz on Monday November 16, 2015 @12:50PM (#10575822)
Heat flow is akin to 'information' flow in a neural net because the weight of a node is the sum/average of the weights of all nodes it's connected to just like the heat of a node is the sum/average of the weights of all the nodes its connected to.