Comparison of artificial neural networks and logistic regression as potential methods for predicting weed populations on dryland chickpea and winter wheat fields of Kurdistan province, Iran
The dominant weeds at dryland chickpea and winter wheat fields in Kurdistan province of Iran were Convolvulus arvensis and Tragopogon graminifolius. ANN could develop the best suited models for prediction all dominant weeds compared to LR models. Sensitivity analysis revealed that altitude and rainfall factors were the most significant parameters in the prediction of weed presence, which are often recognized as sensitive factors. This survey reveals the potential of ANN as a promising tool for the study of weed population dynamics and identifying relationship between dependent. We will be able to make more accurate management decisions based on the implementation of weed distribution maps as well as weed presence prediction using ANN models.