EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Improving Rice Production Prediction Using XGBoost with Feature Engineering of Weather and Land Characteristics

Sartika Lina Mulani Sitio, Syafril

Computer Journal · 01.02.2027 · 5 (1) · 99-107

DOI: 10.58477/cj.v5i1.569 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Computer Journal, online erschienen am 01.02.2027. Verfasst von Sartika Lina Mulani Sitio und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Rice production prediction is needed to support the use of weather information and land characteristics in estimating production yields more measurably. This study aims to develop a rice production prediction model using XGBoost by utilizing weather factors and land characteristics engineering features. The dataset used consisted of 243 observations with 84 initial variables, which were divided into training and testing data with an 80:20 ratio. Feature engineering produced derivative features in the form of aggregation of rainfall, temperature, humidity, solar radiation, as well as nutrient ratios and soil texture. Furthermore, feature selection using XGBoost selected the 30 most relevant features to build an improved model. Hyperparameter optimization was performed using Optuna, while model performance was evaluated using MAE, RMSE, and R². The test results showed that the baseline model obtained an MAE of 0.9049, an RMSE of 1.1576, and an R² of 0.5843. The improved model produced an MAE of 0.9717, an RMSE of 1.2729, and an R² of 0.4974, indicating no performance improvement compared to the baseline. SHAP analysis showed that temp_Dec, N_to_K_Ratio, bulk density, and TMIN_Mean were the features that contributed most to the prediction.

Zitieren

APA

Sitio, S. L. M., & Syafril (2027). Improving Rice Production Prediction Using XGBoost with Feature Engineering of Weather and Land Characteristics. Computer Journal, 5(1), 99-107. https://doi.org/10.58477/cj.v5i1.569

Vancouver

Sitio SLM, Syafril. Improving Rice Production Prediction Using XGBoost with Feature Engineering of Weather and Land Characteristics. Computer Journal. 2027;5(1):99-107. doi:10.58477/cj.v5i1.569

BibTeX
@article{Sitio2027,
  author = {Sitio, Sartika Lina Mulani and Syafril},
  title = {Improving Rice Production Prediction Using XGBoost with Feature Engineering of Weather and Land Characteristics},
  journal = {Computer Journal},
  year = {2027},
  volume = {5},
  number = {1},
  pages = {99--107},
  doi = {10.58477/cj.v5i1.569}
}
Kostenlos starten →