Below you will find a rather abstract account of Estimation and Decision Making. This website does not exist for such dry material, instead it is for realworld applications such as a Justice System. As explained on the website for that specific application, the biggest problem with decision making and estimation is error covariance. Another important application of the material presented below is in bipartite matching. In this case the problem is to estimate the weights applied to each possible connection, so the the resulting solution will be optimal. When is this important? A very specific application is in finding matching pairs of people, to form good interpersonal relationships.
Alright, now for the dry material of the original version of this page.
Decision Estimation has two components. One is the actual estimation of some value. The other is making a decision by comparing that value with some choice function. Often that choice function is a very simple one based on a single cutoff point. Values below that point represent a No, values above it represent a Yes. This may all be done unconsciously by a person, and is certainly an oversimplification, but that’s the basic idea.
This has obvious applications in business and science, but this website was created with social technology in mind, especially that of the justice system. The basic principles apply to all applications however, and may be of great interest to those who write business software. Even when considered for application in social technology, analogies from business are useful.
A key technique in decision estimation is error covariance minimization. That is just as important in the creation of a successful focus group to predict product acceptance in business as it is in assembling a group of colleagues or finding a good interpersonal relationship with social technology.
This and other methods of decision estimation will be discussed on this site. It is still new, with little content, but please check back soon.