Artificial Neural Network Based Apple Yield Prediction Using Morphological Characters

The yield of the crop is a complex function number dependent traits, which makes prediction statistically difficult task. A work on using morphological characters already exists in literature. Most used statistical techniques such as linear regression and models, assume relationship between traits; actual practice, seldom achieved. With advancement field machine learning techniques, these methods can provide viable alternative for dealing with nonlinear relationships prediction. Globally, apples are most consumed fruit. In this paper, attempts have been made to predict apple traits. PCA was selection significant variables. These variables were later input ANN model different hidden layers predicting yield. predictive performance evaluated standard tests. Sensitivity analysis performed find out individual effects each character study contributes better understanding