Foliar tissue analysis is utilized to diagnose a crop's nutrient status. For most floriculture crops a survey approach of a small population of plants (n= <25) of healthy appearing plants are used to establish sufficient nutrient standards. While this historical approach offers a baseline for the wide variety of floriculture crops there is a need for scientifically based ranges similar to those available in agronomic crops. For fast-maturing crops, utilizing foliar tissue analysis and correctly interpreting the results is critical in making fertility adjustments when problems arise. Foliar tissue analysis results of petunia (Petunia hybrida) were compiled from a variety of diagnostic and research institutions to account for variations of growing environments and classified into five ranges (deficient, low, sufficient, high, and excessive). To aid in foliar tissue analysis interpretation machine learning models were evaluated for accurate percent correct classification (PCC) into the sample's respective nutrient classification. Four separate machine learning algorithms were performed to analyze the data set including sequential minimal optimization (SMO) of support vector machines (SVMs) and multilayer perceptron (MLP) artificial neural network (ANN), and two decision tree models J48 and Random Forest (RF). Machine learning algorithms were compared to identify significant model nutrients based on a complete foliar tissue analysis report of 11 elements for the observations. The performance of both machine learning algorithms SMO and MLP were determined using PCC and during the cross-validation. By evaluating the foliar tissue concentration dataset of multiple species by 10-fold and 66% split cross-validations, the incorporation of five elements of ranked based on Shannon Entropy (Information Gain) was able to correctly classify tissue concentrations into one of five foliar nutrient classifications greater than traditional statistics. This information provides additional insight as to how examining nutrient relationships can assist in identifying fertility problems and classifying nutrient ranges.
