Linear Regression and Markov Chain Model for the Arabian Horse Registry.

摘要

This thesis is based on a project done for the Arabian Horse Registry, which needed to forecast its future registration of purebred Arabian horses. A linear regression model was utilized to forecast future registration of purebred Arabian horses. This forecast will enable the Registry to identify future revenues for its operation. Additionally, the Registry wants to increase Pure-bred Arabian's share of the horse market. A Markov chain model was developed to forecast the future horse market based on an independent survey conducted by Elrick Lavidge Marketing Research Inc. This model forecasts which horse breeds will increase and decrease within the horse market. The Arabian Horse Registry has registered fewer foals every year since 1986. Using regression analysis, we forecast this trend to continue for the next three years.