[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference heron::euro_swas_ai

Title:Europe-Swas-Artificial-Intelligence
Moderator:HERON::BUCHANAN
Created:Fri Jun 03 1988
Last Modified:Thu Aug 04 1994
Last Successful Update:Fri Jun 06 1997
Number of topics:442
Total number of notes:1429

75.0. "Reference WANTED" by CHEOPE::EANDI (AI is better than none) Tue Jan 31 1989 16:00

    
    	I have two (different) customers interested in applying
    	AI techniques to SALES FORECAST domain.  As usual their
    	first request has been on references  (AI and not-AI
    	similar projects going on in the world) ......
        
    	I'll meet them next week - but up to now don't get any
    	significant result in my search.
    
    	Can anybody help me?                                        
        .... Any hints will be strongly appreciated!
    	
    	Ciao,
    	Antonella
    
    	
    	
    
    
    
T.RTitleUserPersonal
Name
DateLines
75.1A pointerHERON::ROACHTANSTAAFL !Tue Jan 31 1989 18:567
    One of the example Aise demo's shown at the last AI Forum was a
    forcasting system for VT220 shipments from one of our manufacturing
    plants. Not sure, but I think that there are some parallel's between
    the # of Vt220's we sell and the # we ship! Why not drop George
    McGregor a line for further details?
    
    Pat
75.2ai in stats relevant?MUNSBE::BRITTAINPeter, EDU Munich @UFH *773-2069Wed Feb 01 1989 09:569
    
    Antonella,
    	Is there any overlap between sales forecasting and market 
    research ?  I have had no luck with reference SYSTEMS there either,
    but see notes 69.* for AI in statistics.
    
    ciao,
    Pete.
    
75.3YIPPEE::BOYLEThu Feb 02 1989 11:044
    The forecasting application referred to in .1 was written by Emmanuel
    Brandt, in EAITC. You should call him for details (828-5595)
    
    Frank.
75.4MOCHERON::BUCHANANAndrew @vbo/dtn8285805/ARES,HERONThu Feb 09 1989 16:5918
	One of the facets of MOC (the Manufacturing Operating Consultant) is
to translate from raw dollars sales forecasts (the "Red Book") into
"Top 80" Demand estimates for a hundred or so individual line items.

	The idea is that the dollars estimates come in from the field, and are
comined with historical sales information to determine what the likely demand
is for which CPUS, peripherals etc.	The current demo, which is set in the
sanitized domain of biscuit manufacture, slightly obscures this process, 
because any one biscuit type always has the same ingredients, whereas the
Digital world does not contain a finite set of fixed configurations, each with
a fixed set of ingredients.

	The spiel is that the "redbook process" used to take 3 weeks to perform
(there's a lot of clerical mucking about involved:  I know, I studied this
area in a previous incarnation) and that now it takes 3 minutes.   So this is
a really relevant application.

Andrew.