The Naked Scientists

The Naked Scientists Forum

Author Topic: Can anyone recommend a book on Network Theory?  (Read 7603 times)

Offline Ben Aldhouse

  • Full Member
  • ***
  • Posts: 65
    • View Profile
    • Random and Twisted
Can anyone recommend a book on Network Theory?
« on: 30/09/2007 11:02:49 »
Does anyone here have any recommendations for an easy-to-read book on Network Theory?
« Last Edit: 28/04/2008 08:44:31 by chris »


 

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • Posts: 12656
  • Thanked: 3 times
  • A stitch in time would have confused Einstein.
    • View Profile
Re: Can anyone recommend a book on Network Theory?
« Reply #1 on: 30/09/2007 16:29:21 »
Do you mean computer networks?
 

Offline dkv

  • Sr. Member
  • ****
  • Posts: 299
    • View Profile
Re: Can anyone recommend a book on Network Theory?
« Reply #2 on: 30/09/2007 16:47:33 »
There are various types of networks.
1.Digital Network 
2.Biological Network (TSP Network.each node computes pleasure)
3.Fuzzy-Neural Netwrok(partially digital)
Which network you wish to study?

 

Offline Ben Aldhouse

  • Full Member
  • ***
  • Posts: 65
    • View Profile
    • Random and Twisted
Re: Can anyone recommend a book on Network Theory?
« Reply #3 on: 30/09/2007 22:07:17 »
I heard a recommendation* for 'Linked: The New Science of Networks' on TWiT and wondered if anyone had read it, or anything similar that they would recommend.

*Well it was for an audio version of the book and as a part of an advert for their sponsor.
 

Offline dkv

  • Sr. Member
  • ****
  • Posts: 299
    • View Profile
Re: Can anyone recommend a book on Network Theory?
« Reply #4 on: 01/10/2007 07:33:53 »
Wonderful reading it will be I guess.
It defines network as "where everything is connected to everything else"
nd encourages on to think networks.
But I dont think it matters whether everything is connected to everything or not because connectedness doesnt gurantee non-probabilistic understanding.
The causes will still remain fundamentally random.
If events can not be expalined without using randomness then this framework as good as no framework.
If every node has random component then how will it help to understand everything under Networks.
The replication model of gene along with randomly disputed natural selection is another Network which doesnt explain anything.
 

Offline Ben Aldhouse

  • Full Member
  • ***
  • Posts: 65
    • View Profile
    • Random and Twisted
Re: Can anyone recommend a book on Network Theory?
« Reply #5 on: 01/10/2007 10:16:33 »
Heh, I'm sorry, I'm out of my depth here... that's partly why I want to find to good, easy to digest literature.

I like the idea of looking at the concept of 'randomness'. Perhaps there is no such thing as 'randomness' - just a threshold below which the subtlety and complexity of interactions are beyond what we can realisticly cope with.
 

Offline Ben Aldhouse

  • Full Member
  • ***
  • Posts: 65
    • View Profile
    • Random and Twisted
Re: Can anyone recommend a book on Network Theory?
« Reply #6 on: 01/10/2007 10:22:02 »
Do you mean computer networks?

All networks! Mwa ha ha ha!
 

Offline dkv

  • Sr. Member
  • ****
  • Posts: 299
    • View Profile
Re: Can anyone recommend a book on Network Theory?
« Reply #7 on: 01/10/2007 10:50:03 »
Writing it all again:
If the root causes are random then the consequences or results are randomly distributed over time and space.
It the root causes are non-random in nature then the results will not be evenly distributed and one can detect the anomaly in randomness.
Therefore if the Network works on the basis of random principles then we will find random event space.
But purely random undertanding doesnt work.... at any level.
We begin with the principles of randomeness ... assign some purpose and then find that observations do not match with the theory. There is an overshoot or the undershoot.
I call this anomaly in randomeness.
Anomaly of randomness gives the branch of probability a new dimension. It changes fortune for no reason.... A small chance can manifest for no apparent reason. This anomaly is found everywhere... The Universe I believe came into existence due to this anomaly.
Example: 1 company manages to survive out of 100.
If industries came out of pure random field then there would have 0.5 success rate.That is
(number of Companies Survived)/(Total attempts to establish compnaies)
There are two choices for any given attempt : Either it survives or it doesnt...
BUT this doesnt happen.
Another example: Wealth is unevenly distributed. 1% people hold 90% of wealth. If the background was random then one would have expected near linear distribution of wealth.
HOW DO we explain this?
Anomaly in randomness.

========================
Considering the Life networks:
If we consider that the desires itself manifests randomly then everyone should have desired every other thing once in a life time. (Hence the random natural selection would have led all possible states)

BUt this is not the case.As there are infinitely many possible biological states and only few have survived.(similar to the wealth distribution scenario)
Why?
I consider Life network as an instance of TSP Network.
TSP - Towards Sustainable Purpose
When TSP used for Life it becomes - towards sustainable pleasure.

Life desires one and only thing and that is pleasure.
There is no choice so the question of randomness doesnt arise.
This behaviour is found in TSP network.
TSP networks gives a sense of direction to each node.
Whereas the Replication model of genes doesnt give any direction to the nodes.
By assigning Purpose to everything I remove the anomaly as all questions and answers are with respect to TSP.
===================================
Some(digital or fuzzy) networks have locally define purpose.These purposes are space and time variant and need frequent corrections. Therefore are not independent of Humans.
And humans without knowledge produce error and these errors get amplified and we fail to explain.
In other words the individual nodes never undertood the nature of things. And therefore they can expereince overshoot or undershoot frequently. And hence the network fails to explain cause and effect in proper way.
One can still understand by introducing random errors.
AND THIS is what any non-TSP network does.
It replaces intrinsic-purpose with randomness to account for the result due unaligned nodes of the network.
BUT how do we explain the anomaly in randomness?
===================================

In my opinion every networks tends to be a TSP type.That is every node is congruent to the purpose of Network itself.










« Last Edit: 01/10/2007 11:47:19 by dkv »
 

Offline _Stefan_

  • Hero Member
  • *****
  • Posts: 814
    • View Profile
    • My Photobucket Album
Re: Can anyone recommend a book on Network Theory?
« Reply #8 on: 01/10/2007 14:19:47 »
Please DKV, stop talking about TSP. You have no evidence to support it, the concept is absurd, and a logical analysis of current information will show you that TSP is wrong.
 

Offline dkv

  • Sr. Member
  • ****
  • Posts: 299
    • View Profile
Re: Can anyone recommend a book on Network Theory?
« Reply #9 on: 01/10/2007 14:37:31 »
You think this absurd !!
Read "What is true and false?" or "Quantum wave funtion collpase"  or "Earth Vs. Venus"

I have good evidence it support(and I have repeated it hundred times) :
Condoms
Soft Genetials
Homosexuality
Octopus sexual behaviour(both homosexual and sexual)
Relative less presence of Asexual species or Self pollinating ones.
You do not wish to see and that is can be explained due to your association with Group.
I blame Selfish Reasoning. Not you. Because it is natural.
 

Offline engrByDayPianstByNight

  • Sr. Member
  • ****
  • Posts: 105
    • View Profile
Re: Can anyone recommend a book on Network Theory?
« Reply #10 on: 09/10/2007 16:18:14 »
Hello Ben,

   The term "network theory" is more related to a branch of mathematics (the graph theory). I assume that's not what you want. You're more interested in knowing how "communications networks" work, right? In that case, as some of the folks on this post pointed out, there are a variety of networks, each with its own unique characteristics that may not be generalized so easily.

   Since each type of communications networks has pretty much volumes of books written about it, it's probably a good idea that for the basics, you can get some crash course from the web. I can give you a wiki link on the computer networks:

   http://en.wikipedia.org/wiki/Computer_network

Hope this helps.



 
 

Offline Ben Aldhouse

  • Full Member
  • ***
  • Posts: 65
    • View Profile
    • Random and Twisted
Re: Can anyone recommend a book on Network Theory?
« Reply #11 on: 13/10/2007 11:50:41 »
Thanks engrByDay. You have provided me with a new and interesting concept to look up and read about - 'Graph Theory'. I may have come across this before but forgotten. I have a terrible memory. I have a basic knowledge of artificial neural networks (Adalines and Perceptrons) and am very keen to improve my understanding of more sophisticated techniques.

I am interested in the way everything connects to everything else. That might sound vague - but why not have that as an interest? In a sense, then, I could say that everything I have ever experienced has been an inspiration. I have particularly enjoyed television programs like James Buerk's Connections (Ned [Non existant deity]! That was a long time ago!) and Douglas Adams' book 'Dirk Gently's Holistic Detective Agency' (Both of these I remember almost nothing about other than very much liking the ideas behind them).

Perhaps there could be non-centralised and distributed 'Network of Trust' ways of accomplishing some tasks which could make politics, legal systems etc. more open and accountable? I would love to hear if anyone has heard of or been involved into looking into anything like this.
 

Offline engrByDayPianstByNight

  • Sr. Member
  • ****
  • Posts: 105
    • View Profile
Re: Can anyone recommend a book on Network Theory?
« Reply #12 on: 16/10/2007 03:07:53 »
Ben, if I remember right from somewhere, there is the "six degrees of separation," which was a 1970s social experiment carried out to show that basically (from wikipedia):

"if a person is one "step" away from each person he or she knows and two "steps" away from each person who is known by one of the people he or she knows, then everyone is no more than six "steps" away from each person on Earth."

I don't know much about this experiment or social science in general, but to translate it in graph-theoretical terms, you might say that each vertex (i.e., a person) can reach any other vertex in the whole network (i.e, Earth) through a path of no more than six edges (or hops).

    I know the above doesn't answer your question about whether a distributed network in social sciences may lead to more transparent/accountable legal system/political apparatus. I can only offer an example from an engineering point of view. However, you could probably start with reading some material on the Six Degrees of Separation and see what else that experiment revealed other than the connectedness between individuals (e.g., does more connectedness lead to more honesty between people, etc.?). That might get you a step closer to what you're looking for.
 

Offline Ben Aldhouse

  • Full Member
  • ***
  • Posts: 65
    • View Profile
    • Random and Twisted
Re: Can anyone recommend a book on Network Theory?
« Reply #13 on: 22/10/2007 13:28:20 »
There was a program on telly fairly recently in which a journalist attempted to create a demonstration of the 'six degrees of separation' by finding someone who would seem to be very remote (a Mongolian herdsman, as it happened) and then finding how many links there were between herself and the remote person. I think she found a shortest chain of seven or eight links. It seems to me that she would have had quite a hard time defining what a 'link' was, though. Relationships between people tend to be very complicated. However, I think that it as an interesting idea which reflects how people are, in some ways, becoming less remote from each other.

There was also a film - I think it might have been called 'Six Degrees of Separation' about a con man who claimed to be Sidney Poitiers' son. I don't remember much about it, though. Did anyone else see this film?
 

Offline engrByDayPianstByNight

  • Sr. Member
  • ****
  • Posts: 105
    • View Profile
Re: Can anyone recommend a book on Network Theory?
« Reply #14 on: 23/10/2007 23:06:09 »
Yes, I've seen that movie, with Will Smith, Donald Sutherland, and Stockard Channing. It was a little deep for me, though.
 

Offline Ben Aldhouse

  • Full Member
  • ***
  • Posts: 65
    • View Profile
    • Random and Twisted
Re: Can anyone recommend a book on Network Theory?
« Reply #15 on: 24/10/2007 20:59:28 »
Ah, Stockard Channing... We go together... like do wap di wap de bop.

A few years back, I seem to remember there were a couple of guys who had a small amount of fame from being able to link just about any two films together with just a few links. They seemed to know an awful lot about films. Their fame probably waned just as IMDB was waxing.
 

Offline nikomaster

  • Jr. Member
  • **
  • Posts: 16
  • A new world, new technology science for all
    • View Profile
Re: Can anyone recommend a book on Network Theory?
« Reply #16 on: 05/11/2007 01:54:33 »
A network is just a set of patterns intercontected by a core that manages that network, all this patterns work together sharing information , building the structure, and doing an specific task.

 

Offline Ben Aldhouse

  • Full Member
  • ***
  • Posts: 65
    • View Profile
    • Random and Twisted
Re: Can anyone recommend a book on Network Theory?
« Reply #17 on: 19/04/2008 12:02:34 »
I have just read the 'Are We Doomed?' articles in a recent 'New Scientist'. It is interesting that the second article posits that civilization is in danger from becoming over complicated and the shift civilization is making from typically hierarchically structured to something more network-like isn't necessarily helping matters*. 

I'm sorry I've been away from this forum for a while. However, having read those articles has given  me the incentive to bump up this thread and ask again if anyone has read anything interesting (and fairly accessible!) about networks recently...

Thanks,

Ben.

* I liked the idea they were describing civilization in terms of how it works as a network. I think that there were some comforting points made, too - like how American power companies had learned to build redundancy in to the grid after the massive failures of 2003.
« Last Edit: 19/04/2008 12:07:20 by Ben Aldhouse »
 

The Naked Scientists Forum

Re: Can anyone recommend a book on Network Theory?
« Reply #17 on: 19/04/2008 12:02:34 »

 

SMF 2.0.10 | SMF © 2015, Simple Machines
SMFAds for Free Forums