Re: independence and multiple viewpoints

[Jan Kreft]

Hi,

On Tue, 2 May 2000 address@hidden wrote:

> jan kreft wrote
> >so the problem is lack of independence. If two versions would be
> >independent, output agreement would be sufficient, right?
> >
> >Of course, true independence can't be achieved, but it is the goal.
> >
> >And I wouldn't know of anything better than comparing two versions that
> >are as independent as possible.
> 
> and miles parker wrote:
> >Please someone correct or confirm, but isn't the point that two completely
> >independent implementations _will_ tend to recreate errors in the same
> >places, thus overturning the assumption that multiple versions of a model
> >validate the model? This conclusion challenges some assumptions, but makes a
> >lot of sense to me.
> 
> miles has the gist of my rememberance about profesor leveson's work right
> (i've heard her speak about it, but not read the book either) - independence
> of modeling effort is not the same as independence of modeling errors

Hm, I don't get this. Why should _independent_ implementations tend to
recreate errors in the same places. This is against my intuition, and I
can't think of reasons why this should be so. Maybe we mean different
things when we talk about independent. I mean different algorithmic
solution to the same problem. If you go from A to B via two completely
different routes, you are unlikely to be run over by the same bus, not?

Cheers, Jan.
 
> i also agree with jan that multiple different viewpoints can help in
> validation of complex systems - we've advocated using multiple models in all
> of our modeling applications (i've also shown it to be helpful in some
> communication protocol simulation and verification modeling long ago), and
> raised it to a principle in our knowledge-based integration infrastructure -
> NO one model or modeling method can answer all important questions about a
> complex system (which then leads to a requirement for developing techniques to
> integrate information across different models and modeling methods)




                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================