predictr@bellatlantic.net wrote:> > Will Dwinnell wrote: > > "Here is a simple fuzzy logic system, borrowed from an example given in > "The Fuzzy Systems Handbook", by Earl Cox. The problem is to establish > the price of a product. The fuzzy system has 4 rules: > > 1. The price should be high > 2. The price must be low > 3. The price must be around 2 times cost > 4. If the competition price is not very high, then the price should be > near the competition price >
***************** Dear Will I worked on>>Fuzzy Pricing<< for ~one year back circa 1995. The subject is a little more complicated than what Earl Cox has suggested. However, Earl is in the spirit and his system should produce reasonable answers. I am personally surprised that Fuzzy Pricing has not caught on. At the time, there was difficulty in Management to believe in the system and maybe Radford Neal comments address some of the concerns. There is no doubt some model (Game Model) must be made wrt the competitors and what they are likely doing and how they may be pricing. Pricing has little to do with cost except as Earl Cox is stating some reasonable assumption must be made about the profit region interval. The comment made about no attempt to model the (Nonlinear) Demand Elasticity of the market and if you will the individual ones seen outbound by each company in that market for that product by Radford Neal is well taken. Demand models should be made and used as constraints. Finally I would like to add that I arrived at a Fuzzy Price Interval from a method that maximized the Fuzzy Expectation for my company. FYI, I used definitions like Ultra Low Price Low Price High Price Ultra High Price. I think this discussion in the NG is healthy and both Earl and Radford have relevant points. I still believe pricing is fuzzy and can be mechanized as a>>strong aid<< to actually setting the price in a standard product to the market or to competitive bidding in which was my arena at that time. Paul V. Birke, P. Eng. NN Researcher Guelph ON CANADA