Estimating Quantity of Date Yield Using Soil Properties by Regression and Artificial Neural Network

Models for predicting coffee yield from chemical characteristics of soil and leaves using machine learning

“…The selection of these attributes is a very important task for predicting yield, as highlighted in previous works. [18][19][20] Thus, some variants were selected for the study, such as pruning times, data collection year (2017, 2018, and 2017 + 2018), all chemical attributes of soil and leaf (called 'Leaf/Soil'), soil and leaf attributes that have the highest Pearson correlation with yield, S (leaf), MO (soil), S (soil), Zn (leaf), Ca (leaf), K (leaf), T (soil), P (leaf), B (soil) and C (soil); and, finally, only the chemical attributes of leaf with the highest Pearson correlation, namely: S (leaf), Zn (leaf), Ca (leaf), K (leaf), and P (leaf ).…”

“…Regression models were employed to predict coffee yield using the following algorithms: random forest (RF), K‐nearest neighbors (KNN), linear regression, decision tree, support vector regression (SVR), and AdaBoost. These algorithms were also used in the literature, utilizing soil characteristics as attributes for model learning 18‐20 . The use of the flowchart (Fig.…”

“…These algorithms were also used in the literature, utilizing soil characteristics as attributes for model learning. [18][19][20] The use of the flowchart (Fig. 1) to predict the yield data of the studied crops was based on an analysis of the chemical attributes of the soil and leaves in two consecutive harvests.…”

“…Seyedmohammadi et al 18 used regression tree, k-nearest neighbors, and support vector machines to predict yield and manage the use of natural resources sustainably in pistachio production. In another study, Eskandari et al 19 presented a methodology to estimate date production using soil characteristics and machine learning methods, multivariate regression, and artificial neural networks. In research 20 that employed advanced remote sensing methods, the authors used support vector machines improved by metaheuristics to predict the amount of water in the soil.…”

“…used regression tree, k‐nearest neighbors, and support vector machines to predict yield and manage the use of natural resources sustainably in pistachio production. In another study, Eskandari et al 19 . presented a methodology to estimate date production using soil characteristics and machine learning methods, multivariate regression, and artificial neural networks.…”