A decision analysis and expert system model was applied to the problem
of individual nutrient selection for cucumber growth in a controlled
environment. The ability to use a decision tree with probabilities tog
ether with intuitive modeling for the derivation of a plant growth fun
ction through the use of expert systems provided a flexible and adapta
ble model. The model was assembled using a decision tree as the core w
ith three rulebases providing inputs into the tree. The decision tree
was comprised of a decision node (which nutrients?), a chance node (pr
obabilities of solar irradiance ranges) and an outcome (plant growth f
unction). The rulebase for the plant growth function was derived by gr
owing a cucumber crop in a greenhouse at the Ohio Agricultural Researc
h and Development Center, Wooster, Ohio. The expected value of each re
cipe was determined by computing the cross-products of the probabiliti
es of irradiance levels and the plant growth function.