[Prev][Next][Index][Thread]

Re: NEON- Patented Processing System



Dear Kenny and all, 

when setting up the variables for a statistical test procedure, 
remember how many variables we have and how they're connected. 

In my opinion this is not only a Ph.D. thesis work for a physicist, 
it will be a life work. 

The problem is you cannot vary the parameters you need for a test 
according to the requirements of a given job. 
Who will pay for the test tubes you all have to make and you can never 
sell, when covering i.e. extreme parameters? 

I keep a complete record of any important parameter of each tube I EVER made.
It's impossible for me to make a statistical scheme for classification. 

So you may think about the amount of tests and data needed prior to 
start those experiments. I won't be lucky making test tubes all my life.....

Marcus

 
> 
> I agree with John's sentiments. I'd add what I assume you know, 
> that the statistical model or model(s) you choose are extremely 
> important.  It looks like you have an interesting procedure.  The 
> trick is to make sure that the data you see is really telling you 
> what you think it is saying.  
> 
> Are you working with a model at this point that tests statistical 
> significance?  In crude terms, this filters out the level at which a 
> portion of the results you see may be the by-product of chance.
> 
> Anyway, I'm very impressed. Good luck.
> 
> Kenny
> 
> 
> 
> 
> > the potential for ground-breaking research.  I only hope the
> > scientific method is adhered to.
> > 
> >     -John  
> > 
> Kenny Greenberg --  Neon - Scenic and Environmental Art  
>                     Internet Site Consultant and Author   
> KRYPTON NEON 34-43 Vernon Blvd Long Island City, NY 11106
>          Phone: 718-728-4450  Fax: 718-728-7206  
> http://www.neonshop.com  -  The Internet's Neon Shop
> http://www.licweb.com  - The Long Island City Web
> 


References: