Article: Forecasting Occupancy Levels: A Problem for Management Forecasters Journal: Journal of Business and Management (JBM) 2006 Vol.12 No.1 pp.59 - 69 Abstract: Accurately forecasting occupancy levels is an essential planning ingredient in the process of developing a forecast of total annual revenue in managing an institution of higher education. Literature suggests that applied time series analysis is preferable to causal models for this application. The problem is that time series methods may not be optimal for forecasting total occupancy level due to the presence of measurement error in the historical occupancy data. Song and Chissom (1993b, 1994) developed a time series method based on Euzzy Logic Set Theory (Zadeh, 1965), which they report to be robust in the presence of this type of uncertainty. We compare six general time series methods and the fuzzy method of Song and Chissom across two direct and two derived forecasting strategies. We also compare six time series methods and the 'fuzzy' across direct and composite forecasting. We evaluate these combinations of method and strategy based on the criteria of accuracy and economy. The results indicate that the fuzzy time series method in conjunction with a direct forecasting strategy is preferable to all other combinations of method and strategy evaluated for forecasting each semesters total occupancy level. Inderscience Publishers - linking academia, business and industry through research

Title: Forecasting Occupancy Levels: A Problem for Management Forecasters

Authors: Jeffrey E. Jarrett; Jeffrey S. Plouffe

Addresses: Author address listing can be found in the "About the Authors" section at the end of the article.

Abstract: Accurately forecasting occupancy levels is an essential planning ingredient in the process of developing a forecast of total annual revenue in managing an institution of higher education. Literature suggests that applied time series analysis is preferable to causal models for this application. The problem is that time series methods may not be optimal for forecasting total occupancy level due to the presence of measurement error in the historical occupancy data. Song and Chissom (1993b, 1994) developed a time series method based on Euzzy Logic Set Theory (Zadeh, 1965), which they report to be robust in the presence of this type of uncertainty. We compare six general time series methods and the fuzzy method of Song and Chissom across two direct and two derived forecasting strategies. We also compare six time series methods and the 'fuzzy' across direct and composite forecasting. We evaluate these combinations of method and strategy based on the criteria of accuracy and economy. The results indicate that the fuzzy time series method in conjunction with a direct forecasting strategy is preferable to all other combinations of method and strategy evaluated for forecasting each semesters total occupancy level.

Keywords: Occupancy forecasting; time series analysis; fuzzy logic; revenue prediction; management forecasting; higher education planning.

DOI: 10.1504/JBM.2006.141140

Journal of Business and Management, 2006 Vol.12 No.1, pp.59 - 69

Published online: 05 Sep 2024 *

Title: Forecasting Occupancy Levels: A Problem for Management Forecasters

Keep up-to-date