An intelligent system based agriculture microsimulation model with bootstrapping validation

Advanced level models are essential in dealing with big data environments. Recent developments in smart technologies now allow environmental characteristics like climate, land cover and topography to be measured at fine spatial resolutions. But, reliable small area level data with enough comprehensive information for measuring human socioeconomic activities is often unavailable because they contain confidential or proprietary information. For examples, spatially-disaggregated and explicit data depicting the distribution of livestock, crops, and invasive animals is needed to inform potential agricultural production increases and to manage the risks to food safety and biosecurity. The lack of a small area datasets often negatively impacts the ability of agencies to manage serious health risks. The attributes of individual farms, e.g. their distances to other farms, the size of the population, and invasive animals data is needed to parameterize spatial epidemiological models. Spatial microsimulation models are powerful tools to simulate the characteristics of populations, and model the complex interactions of individuals’ activities at the finest spatial resolutions. This paper provides a new intelligent system based algorithmic methodology for a tractable class of spatial microsimulation models for agricultural industries. It also demonstrates bootstrapping reliability measures for the small area agriculture estimates. Empirical findings reveal that the proposed intelligent system based modelling technique is performing better than the traditional approaches to spatial microsimulation modelling. The bootstrapping validation results show an
enhanced reliability to the small area agricultural estimates for regions with smaller populations.