A Decision Support Tool for Buying Farm Tractors, Based on Predictive Analytics

Data science can help farmers when making a decision about tractor purchase. Buying represents big investment for farmers, and price is one of their main concerns. This study presents the development web-based support tool (DST) that calculate new second-hand tractors, purpose providing decision-maker some information will lead him to final decision. The makes use different algorithms based on predictive analytics methodologies. dataset has 227 observations tractors 1003 from European countries. During study, prices used were modeled, testing parametric non-parametric regression models with segmentations predictor variables. Non includes trees, vector machines, ensembles Gaussian process, neural networks. In both cases, predicting adjusted R2 higher than 0.99 achieved. developed implemented in DST which fully operative, available Internet, free use.