Skip to content

Instantly share code, notes, and snippets.

@zhouzhuojie
zhouzhuojie / 9.12 Theorems and Experiments
Created September 13, 2013 00:42
9.12 Theorems and Experiments
NB_DEGREE_2: Divide the neighbors into 2 groups
NB_DEGREE_4: Divide the neighbors into 4 groups
Facebook0
![](http://i1336.photobucket.com/albums/o641/00zzj/BTRW/0912/0912_nrmse_degree_4_facebook0_zps05669702.png)
@zhouzhuojie
zhouzhuojie / 9.18 Experiments
Created September 18, 2013 13:57
9.18 Experiments
![](http://i1336.photobucket.com/albums/o641/00zzj/BTRW/0912/0918_nrmse_md5_2_as733_zps2a39b4b1.png)
![](http://i1336.photobucket.com/albums/o641/00zzj/BTRW/0912/0918_nrmse_md5_2_as733_large_zpsf8f11cfe.png)
----
![](http://i1336.photobucket.com/albums/o641/00zzj/BTRW/0912/0918_nrmse_md5_2_facebook0_zps8f02a691.png)
@zhouzhuojie
zhouzhuojie / 9.25 Experiments
Created September 25, 2013 17:06
9.25 Experiments
### as 733, zoomed
![](http://i1336.photobucket.com/albums/o641/00zzj/BTRW/0912/0925_nrmse_md5_oe_as733_zpsae4f876f.png)
![](http://i1336.photobucket.com/albums/o641/00zzj/BTRW/0912/0925_nrmse_md5_oe_as733_zoomed_zps15f7b0b0.png)
### facebook0, zoomed
![](http://i1336.photobucket.com/albums/o641/00zzj/BTRW/0912/0925_nrmse_md5_oe_facebook0_zps8d84cae7.png)
![](http://i1336.photobucket.com/albums/o641/00zzj/BTRW/0912/0925_nrmse_md5_oe_facebook0_zoomed_zps9134f1cd.png)
@zhouzhuojie
zhouzhuojie / 1006 Recent ideas
Created October 7, 2013 02:16
Summary of recent ideas
Recent Ideas
============
Edge-induced Graph
-----------------
* Edge-induced Graph
* Edge -> Node (Mapping edges into nodes)
* From edge to edge, the probability is associated with the node that connecting two edges, i.e. the probability of transfer from edge $e_{ab}$ to $e_{bc}$ is $\frac{1}{k_b-1}$ (non-backtrack to its self)
@zhouzhuojie
zhouzhuojie / 1014 Recent Ideas
Created October 14, 2013 18:52
Summary of recent ideas
Recent Ideas
============
Edge-induced Graph
-----------------
* Edge-induced Graph
* Edge -> Node (Mapping edges into nodes)
* From edge to edge, the probability is associated with the node that connecting two edges, i.e. the probability of transfer from edge $e_{ab}$ to $e_{bc}$ is $\frac{1}{k_b-1}$ (non-backtrack to its self)
@zhouzhuojie
zhouzhuojie / 1028 Recent Ideas
Created October 28, 2013 19:13
Summary of recent ideas
Recent Ideas
============
Edge-induced Graph
-----------------
* Edge-induced Graph
* Edge -> Node (Mapping edges into nodes)
* From edge to edge, the probability is associated with the node that connecting two edges, i.e. the probability of transfer from edge $e_{ab}$ to $e_{bc}$ is $\frac{1}{k_b-1}$ (non-backtrack to its self)
@zhouzhuojie
zhouzhuojie / 1105
Created November 6, 2013 02:56
1105 Experiments and theoretical results
![](http://i1.minus.com/jbfdhzoXmsy314.png)
_**Theorem** Given a graph $G$, and its corresponding edge-induced graph $G^*$. If $A, B, C$ forms an triangle in $G$, and $k_C=2$, then the edge $\{e,f\}$ is not a cross-cutting edge in $G^*$._
_**Proof**_: Prove by contradiction. Let's assume the edge $\{e,f\}$ in $G^*$ is a cross-cutting edge. When $k_C=2$, $P(e,f) = 1/2*(k_C-1) = 1/2$. Given the assumption that $e, f$ are from different partitions, so whenever we try to partition $d$ into $e's$ or $f's$ part, the probability flow connected to $e$ or $f$ must be larger than $1/2$, which makes us can simply drag the one from one partition to another to achieve smaller conductance.
_**Theorem** Given a graph $G$, and its corresponding edge-induced graph $G^*$. If $A, B, C$ forms an triangle in $G$, and $k_A= 2$ and $k_B= 2$, then the edge $\{e,f\}$ is not a cross-cutting edge in $G^*$._
_**Proof**_: According to the first theorem, edge $\{e,d\}$ and $\{d, f\}$ are not cross-cutting, then edge $\{e, f\}$ cannot
@zhouzhuojie
zhouzhuojie / 1108 Experiments
Created November 8, 2013 17:48
1108 Experimental Result
New Experimental Result
==================
![](http://i.minus.com/j7VFX5DMV8Ka8.png)
zoomed:
![](http://i.minus.com/jmVtbAV6ISgGe.png)
@zhouzhuojie
zhouzhuojie / 1113 Experimental Results
Created November 13, 2013 16:49
1113 Experimental Results
Experiments:
* Only consider the 3-degree node rewiring in edge-induced graph
* Not include the non-backtracking part
Observations:
* Previous experiments are invalid due to some implementation bugs
* Only rewiring the 3-degree node is not efficient, sometimes worse than normal SRW.
* The second graph's bias can be interpreted as how do we deal with the probability after removing an edge in the edge-induced graph. We can discuss about it later today.
@zhouzhuojie
zhouzhuojie / 1118 Experimental results
Created November 18, 2013 12:05
1118 Experimental Results
# 3-degree based rewiring only
Facebook0
![](http://i.minus.com/i1tNuCa8jXxMr.png)
Zoomed
![](http://i.minus.com/iDFtGz2W3rXvF.png)