Empirical Analysis of Dataset Size Impact on Classification Performance in Precision Agriculture Using Machine Learning Models
Year: 2025
Type: article
Abstract: This study empirically investigates the relationship between dataset size and classification performance in precision agriculture applications. Seven machine learning models (Decision Tree, Random Forest, Logistic Regression, SVM, Gaussian Naïve Bayes, KNN, and AdaBoost) were evaluated on seven agricultural datasets ranging from 100 to 4,000 samples. Performance was assessed using five metrics: accuracy, precision, recall, F1-score, and ROC-AUC. The methodology involved two phases: initial evaluation using complete datasets, followed by systematic analysis of subdivided datasets to examine performance variation with data volume. Statistical analysis using Pearson correlation coefficients revealed no significant correlation between dataset size and model performance (r = 0.12, p > 0.05). Results indicate that Random Forest and Decision Tree models achieved the highest average performance across datasets (88.48% and 85.37% accuracy, respectively). The findings suggest that dataset quality and problem characteristics have greater influence on classification performance than dataset size alone in precision agriculture applications.
Source: Informatica
Authors: Khadija Lechqar, Mohammed Errais
Institution: University of Hassan II Casablanca
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: diamond
OpenAlex ID: W4417469753
