EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Enhanced Crop Prediction using Adaptive Mutation Swarm Optimized Multilayer Perceptron: A Data-Driven Approach to Maximizing Agricultural Productivity

Glaret Subin P, Sanjay Kumar

International Research Journal of Multidisciplinary Technovation · 30.05.2026 · 404-421

DOI: 10.54392/irjmt26324 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Research Journal of Multidisciplinary Technovation, online erschienen am 30.05.2026. Verfasst von Glaret Subin P und 1 weiteren Autor:innen. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Yield prediction and crop recommendations require precise methodology to increase efficiency of land, agriculture production, and usage of sustainable methods. Current statistical or machine learning methods are inadequate when attempting to model nonlinear associations among climatic environment, soil conditions and suitability of a crop for that particular area. An Adaptive Mutation Swarm Optimization Based Multilayer Perceptron (AMSO-MLP) was developed to produce intelligent crop recommendations according to varying soils and environments. This new algorithm combines particle swarm optimization (PSO) with an adaptive mutation approach through optimization of the multilayer perceptron’s (MLPs’) weights and biases through PSO. As a result of this approach, convergence and prediction are increased in both accuracy and speed, while preventing premature convergence due to local optima. The AMSO-MLP model uses key agricultural components such as nitrogen (N), phosphorus (P), potassium (K), temperature, humidity, precipitation, and soil pH to identify the best crop that can be grown in these conditions. A benchmark data set containing agricultural features and was analyzed for accuracy through a 10-fold cross-validation analysis to demonstrate the effectiveness and strength of the results. Performance metrics included Precision, Recall, F1-Score, RMSE, MAE, and coefficient of determination (R2). The results demonstrate that the AMSO-MLP model is superior to existing neural network-based optimization approaches. The model achieved Precision of 0.989, Recall of 0.987, F1-score of 0.987, RMSE of 0.210, MAE of 0.125, and R² of 0.964, indicating high predictive accuracy and stability. These findings highlight the effectiveness of the proposed hybrid optimization framework for data-driven crop recommendation and precision agriculture, providing a reliable decision-support tool for sustainable agricultural management.

Zitieren

APA

P, G. S., & Kumar, S. (2026). Enhanced Crop Prediction using Adaptive Mutation Swarm Optimized Multilayer Perceptron: A Data-Driven Approach to Maximizing Agricultural Productivity. International Research Journal of Multidisciplinary Technovation, 404-421. https://doi.org/10.54392/irjmt26324

Vancouver

P GS, Kumar S. Enhanced Crop Prediction using Adaptive Mutation Swarm Optimized Multilayer Perceptron: A Data-Driven Approach to Maximizing Agricultural Productivity. International Research Journal of Multidisciplinary Technovation. 2026:404-421. doi:10.54392/irjmt26324

BibTeX
@article{P2026,
  author = {P, Glaret Subin and Kumar, Sanjay},
  title = {Enhanced Crop Prediction using Adaptive Mutation Swarm Optimized Multilayer Perceptron: A Data-Driven Approach to Maximizing Agricultural Productivity},
  journal = {International Research Journal of Multidisciplinary Technovation},
  year = {2026},
  pages = {404--421},
  doi = {10.54392/irjmt26324}
}
Kostenlos starten →