A Neural Network Approach to a Spatial Production System case for milk production

The usual conceptualisation of farmer production system involves three interrelated production systems: a feed production function which inputs are fertiliser, land, weather machinery and labour; a cattle production function based cows, feed - bought or produced - machinery and labour and a conversion production system that generates milk and beef. The aim of this paper is to conceptualise the spatial farm production system in only one system using a neural network's mechanism with the whole set of inputs (fertiliser, land, weather, feed, cows, machinery and labour) and with only one output, milk. We review the concept of agricultural production systems. We systematise some models of agricultural systems. We calibrate a neural network model to the milk production in Terceira Island.