Vector space analysis of the New Zealand Wool Board's auction data
Abstract
A geometric field approach is used to represent technical specification, availability and relative price of the New Zealand Wool Board's sales records. The parameters of price and quantity form scalar fields in the vector space of all wool sold by auction in New Zealand between 1984 and 1988. Dissection is used to reduce the data and define discrete points of aggregated lots with similar properties. Principal component analysis indicates that the supply of New Zealand wool can be adequately described by four independent components.
Current usage and future possible applications for this massive database are outlined and the possible development of an information system which incorporates some aspects of artificial intelligence to provide ready access to this data is discussed.... [Show full abstract]