network exploration
Table of Contents
- DFS - web
- BFS - maze
- - find important node
- metropolis process - find suburb node
Backlinks
a network exploration algorithm. The frequency of a node of degree \(d\) being visit approaches \(\frac{d}{2m}\)
Algorithm: in each step, select a neighbor at random, and move to it. example -
- arrange neighbor in an order
- generate a random number \(i\) in \([0,1)\)
- go to the $\text{floor}(d * i)$th neighbor
network algorithms
Algorithms that operats on network
metropolis process
Algorithm of network exploration that favours node with fewer degree.
Algorithm:
- select a neighbor at random
- set probability of moving to it as \(\frac{d_\test{self}}{d_\text{neighbor}}\) (which is smaller when neighbor have more links, therefore favor node with less links)
- If the probability is larger than 1 (neighbor have fewer links), move to neighbort
- if not, generate a random number in \([0,1)\), if it is larger than the probability, then move to the neighbor/otherwise don’t move