zhouzhuojie

Engedit Manual

Written with Python Flask Framework, if you want to port to PHP or Node.js, just change the templates in app/templates folder.

How to run the website

• SSH Running from Mac or Linux console, use putty if you are on Windows:

zhouzhuojie

Engedit Manual
===========

Written with Python Flask Framework, if you want to port to PHP or Node.js, just change the templates in `app/templates` folder.

How to run the website
------------------------

zhouzhuojie

2norm
![enter image description here][1]

kl-divergence
![enter image description here][4]

errorbars
![enter image description here][2]

errors

zhouzhuojie

New Idea of the fast forward random walk
==================

Definition
----------

The fast forward random walk works as follows: it try to avoid coming back to the former clique.

* Denote the rrandom walk's 2-lag history as $(u_2, v_2), (u_1, v_1)$

zhouzhuojie

# Error Bar plot for distance estimation

### Steps

* Choose an initial node
* Denote other nodes' attributes $d(v)$ as the shortest path's length from $v$ to the initial node (distance to the initial node)
* Estimate $D = \sum\pi(v)d(v)$, where $\pi(v) = \frac{k_v}{2|E|}$
    * Useful and convenient, because we can simply apply the mean function to the sample to get unbiased estimation of $\hat{D}$.

zhouzhuojie

# 1204 2norm avgError KL
 
* 2-Norm and KL-divergence are based on one vertical array's distribution
* One of the average error is estimated using sliding window
 
## 2-Norm
![enter image description here][1]
 
## Average Error  (Non Sliding window)

zhouzhuojie

# 1204 2norm avgError KL
 
* 2-Norm and KL-divergence are based on one vertical array's distribution
* One of the average error is estimated using sliding window
 
## 2-Norm
![enter image description here][1]

## Average Error  (Non Sliding window)
![enter image description here][2]

zhouzhuojie

# 1204 2norm avgError KL

* 2-Norm and KL-divergence are based on one vertical array's distribution
* One of the average error is estimated using sliding window

## 2-Norm
![enter image description here][1]

## Average Error  (Sliding window)
![enter image description here][2]

zhouzhuojie

# Facebook0 dataset
 
## 2Norm-large
![2Norm-large][1]
 
## 2Norm-small
![enter image description here][2]
 
## AvgError-large

zhouzhuojie

# Facebook0 dataset

## 2Norm-large
![2Norm-large][1]

## 2Norm-small
![enter image description here][2]

## AvgError-large