1.Why is beer flavor important to Coors’ profitability?
Because our choices as drinkers depend on various factors such as the mood we’re in, the available venues out there, as well as the occasion, Coors believes if the company could understand the beer flavor based solely on its chemical composition, it would open new avenues in order to create beers that would suit almost every customer’s expectations.
2.What is the objective of the neural network used at Coors?
The fact that the relationship between chemical analysis and beer flavor is not yet clearly understood, but yet substantial data exist of the chemical composition of a beer, as well as the sensory analysis, Coors objective of the neural network is to find a mechanism that link the chemical composition of the beer together with the sensory analysis.
3.Why the results of Coors’ neural network were initially poor and what was done to improve the results?
They were initially poor because of these two major factors: 1.They were primarily concentrating on a single product quality; thus, the variation in the data was pretty low. In addition, the neural network could not extract useful relationship from the data. 2.It was probable that only one subset of the provided inputs would have an impact on the selected beer flavor. Furthermore, according to Wilson and Threapleton (2003), performance of the neural network was affected by “noise” created by inputs that had no impact on flavor.
4.What benefits might Coors derive if the project is successful?
If the project is successful, Coors might gain several competitive advantages by creating beers with flavor that suit a wide range of customers’ expectations. Another benefit that Coors might derive from this project is basically retaining and attracting even more customers for its beers, which would increase its annual revenue tremendously.
5.What modifications would you make to improve the results of beer flavor prediction?
I would probably implement many researches and actual surveys from Coors customers as well as non Coors customers to find what flavor that they are usually in the mood for based on different occasions. In addition, in order to be even more precise on finding flavor that suit our customers, it would be ideal to compute information regarding what beer or what flavor people go for when they are in different moods. Working with more actual data related to customers’ moods, perceptions and taste during different occasions could probably provide better result or prediction for beer flavor that truly satisfy many customers no matter what the circumstances are.