I would like to communicate with someone who has experience developing
ROC modeling curves if possible.  I'm posting here because my research
shows ROC curves are used extensive in radiology.  If I'm in the wrong
newsgroup for this questions please feel free to let me know.

Thank you.

Sent via Deja.com http://www.***.com/
Before you buy.

 > I am interested in developing an ROC model and would like
 > to know whether anyone in this group has done so.  This
 > means designing and writing the code for an ROC curve
 > model rather than using free or commercial software.
 >
 > I would like to communicate with someone who has
 > experience developing ROC modeling curves if possible.
--snip--

Although I have experience in ROC methodology, I don't understand what
you mean by "developing an ROC model."  A new statistical model for ROC
data?  A new parametric functional form for ROC curve-fitting?  A new
criterion for "best-fit" ROC curve?  A new approach for testing the
statistical significance of differences between ROC curve estimates?

  Charles Metz

Over the last 20 years I have worked in a computational environmental
where I have written the code for experimenting with various decision
models and statistical programs in a useful but not so popular
programming language.  I'm very interested in learning how to write a
program for a basic ROC curve model in my (APL)programming language.

I have several articles written by John A. Swets, who as you know is
one of the developers of ROC methodology.  What I meant by "developing
an ROC model" referred to reading and digesting the existing theories
on ROC methodologies to be able to program a ROC model for my own
application.  If I can find very simplistic ROC models in a spreadsheet
application with data or in a printed format where I could trace the
formulas in steps it accelerate my project. I would then be able to
experiment with various detection and threshold alternatives using
my modeling experience and knowledge of my application.

Thank you again for your response.  I will email you with a few
additional comments to make sure you understand what I am trying to do.

Thank you very much for your respond.  It's enlightening to get a
response from someone with your experience in ROC methodology.  I have
been to your site and looked at what is available there.  Additionally,
I plan to read many of the articles and publications you have written
over the years.

Over the last 20 years I have worked in a computational environmental
where I have written the code for experimenting with various decision
models and statistical programs in a useful but not so popular
programming language.  I'm very interested in learning how to write a
program for a basic ROC curve model in my (APL)programming language.

I have several articles written by John A. Swets, who as you know is
one of the developers of ROC methodology.  What I meant by "developing
an ROC model" referred to reading and digesting the existing theories
on ROC methodologies to be able to program a ROC model for my own
application.  If I can find very simplistic ROC models in a spreadsheet
application with data or in a printed format where I could trace the
formulas in steps it accelerate my project. I would then be able to
experiment with various detection and threshold alternatives using my
modeling experience and knowledge of my application.

Thank you again for your response.  I will email you with a few
additional comments to make sure you understand what I am trying to do.

Sent via Deja.com http://www.deja.com/
Before you buy.