ESTIMATING AGRARIAN INCOME EVOLUTION IN HORTICULTURAL FARMS
This paper describes a hybrid method to simulate the evolution of agrarian income (estimated by farm net margin) on a medium-long term basis. The model integrates Monte-Carlo simulation (MCS) and a multi-level fuzzy inference engine. MCS is a well-known forecasting method but ignores the existence of real dependence relationships among covariates. When dependence relationships cannot be algebraically formulated but can be described as rules using expert knowledge, a fuzzy inference engine can interpret them guiding the MCS model. The financial structure of horticultural farms (agrarian year 2000, surveyed in 2003: 348 farms) in Andalusia (southern region in Spain) was used to verify the hybrid model. Farm net margin was forecasted from 2001 to 2013 and the results were evaluated using a cross-validation procedure. Future environmental and market impacts can also be evaluated and this makes our model an interesting proposal as a complementary tool for decision making in the context of agricultural policy.
García-Alonso, C.R. and Pérez-Naranjo, L.M. (2008). ESTIMATING AGRARIAN INCOME EVOLUTION IN HORTICULTURAL FARMS. Acta Hortic. 802, 443-448
Monte Carlo simulation, fuzzy inference, agrarian income, horticultural farms