A matrix
X of size g×g is defined to describe the connectivity inside the group, inside which
an entry xij indicates the existence of a directional link from i to
j. Such a matrix summarizes the whole status of the group and is sufficient to
include all the information needed to describe a group. One example can be:
Here the
power of X is 1, so each non-zero entry indicates a link of length 1. For example,
x12=x15=1, the physical meaning is that node 1 has an out
degree of 2, pointing to node 2 and node 5 once each.
By
taking the nth power (n < g) of X, say n=3, we obtain:
In this
case, n=3 gives all the possible paths of length 3 among the existing nodes.
The intensity of a certain entry gives the number of different paths.
Take
=3 for
example, this means from node 1 to node 3, there exist three paths of length 3,
they each are:
Node 1 – Node 2 – Node 6 – Node 2
|
Node 1 – Node 2 – Node 3 – Node 2
|
Node 1 – Node 5 – Node 6 – Node 2
|
|
|
|
And this
process exhausts all the possibilities.
This
matrix representation method is concise and informative for either visual
observation or statistics analysis. From my point of view, adopting matrices has advantages as follow:
1. Standardized computation.
By applying multiplication to the original matrix, we can easily obtain different level of information without introducing much computation. This way of information processing also allow computers to calculate answers in a much easier way, so that codes and programs can be as concise as possible.
2. Fixed structure of matrices
The format is still concise and matrix remains the same dimension and size as the original one. Such consistency can keep the systems function properly.
3. Easier reuse
This method is not the final goal of a topic. It's just a method used to simply other problems.
In the form of matrices, this calculation method can be easily implemented and encapsulated. In this way, black box reuse is possible, so that users don't necessarily need to know the exact calculation methodologies behind it, instead the numbers say for themselves, and should be enough to explain problems.
4. Can you guys think of more advantages? :)
1. Standardized computation.
By applying multiplication to the original matrix, we can easily obtain different level of information without introducing much computation. This way of information processing also allow computers to calculate answers in a much easier way, so that codes and programs can be as concise as possible.
2. Fixed structure of matrices
The format is still concise and matrix remains the same dimension and size as the original one. Such consistency can keep the systems function properly.
3. Easier reuse
This method is not the final goal of a topic. It's just a method used to simply other problems.
In the form of matrices, this calculation method can be easily implemented and encapsulated. In this way, black box reuse is possible, so that users don't necessarily need to know the exact calculation methodologies behind it, instead the numbers say for themselves, and should be enough to explain problems.
4. Can you guys think of more advantages? :)
Wow, what a fast-moving and good analytic!
ReplyDeleteSimple cuz I wanna put this simple idea into my own words. Anyway it's just a tool, not much tricks. But whoever invented this should get a credit... (废话- -
DeleteNice work ! really helpful :)
ReplyDeleteAs you teaches me last lecture, the numbers in the matrix means the possible paths of length from one node to another node. This blog can definitely explain all process among the Multiplication of matrix.
ReplyDeleteThanks for the appreciation Silu. =)
DeleteI really feel that dealing with matrixes is more fun than dealing with words. Do you agree ? ^^
ReplyDeletecan't agree more
DeleteBesides, I think maths do bring us with many fun.
We write, read and comment, but the maths give us a deep understand of our thinking models
Haha you got it. Especially when these procedures can be simplified using codes and tools. LOL
DeleteGood Blog!Using matrices helps us solve a lot of problems easily, as you said, it not only has the advantage of standardized computation, but also keeps the systems function properly.Thus we can analysis the social network much more efficiently.
ReplyDeleteYes, I think this matrix operation is magical. By multiplication, we can see what is the "distance" between the two nodes and how many ways can be reached from one to another. This reminds me of Markov Chain in my undergraduate lesson. Markov Chain is mainly used to deal with probabilities and measures the probability from one state to another, after certain number of processes. The two operations are alike but they are in 2 areas totally different!
ReplyDelete