Logo Lanfrica
  • Home
  • Atlas
  • Insights
  • Docs
  • Sign in

© 2026 Lanfrica. All rights reserved. All copyrights of the resources shown on the Lanfrica website belong to the original copyright holders, unless explicitly stated otherwise.

Integrating Metaheuristic Optimization with Stochastic Gradient Boosting for Groundwater Potential Prediction in Data-Scarce Arid Environments

Domain:

environment and energygeospatial
Creator:
RacEyyIbrCem
Publisher:
Spr
Host:
Abstract In arid and semi-arid regions, the limited availability of surface water resources makes groundwater a critical source for domestic, agricultural, and industrial use. However, in Djibouti, unplanned and excessive exploitation of groundwater has led to aquifer depletion and seawater intrusion. Therefore, accurate and reliable prediction of groundwater potential represents an urgent need for sustainable water management. In this study, a comprehensive dataset comprising 14 hydrogeological, topographic, climatic, and remote sensing-based variables was developed to model groundwater potential. To predict groundwater potential, the Stochastic Gradient Boosting (SGB) algorithm was applied, while hyperparameter optimization was carried out using three different metaheuristic approaches, namely Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Gravitational Search Algorithm (GSA). Model performance was evaluated using accuracy, precision, recall, F1-score, and ROC/AUC metrics. Results demonstrated that the GSA-SGB model exhibited the most balanced performance on the test set, with 78.6% precision, 73.6% recall, and an F1-score of 0.749. Additionally, the PSO-SGB model achieved the highest classification accuracy (0.8621), while runtime analysis revealed that PSO was the fastest (195 s) and GA the slowest (627 s). ROC analysis indicated that the highest AUC value (0.813) belonged to the GSA-SGB model. SHapley Additive exPlanations analysis further revealed that Normalized Difference Vegetation Index and the sediment transport index were the most influential factors, with low slope and high precipitation conditions significantly enhancing groundwater potential. Overall, the findings highlight that the integration of machine learning with metaheuristic optimization techniques provides a powerful decision-support framework for predicting groundwater potential in Djibouti.

Visit

doi.org

Licenses

https://creativecommons.org/licenses/by/4.0https://creativecommons.org/licenses/by/4.0

Similar

Short Term Power Demand Prediction Using Stochastic Gradient BoostingHybridized light gradient boosting and whale optimization algorithm for diabetes detectionPublic Debt Forecasting in Kenya: Integrating Macroeconomic Variables with Arimax and Extreme Gradient BoostingCrop yield prediction in Ethiopia using gradient boosting regressionA hybrid Machine Learning approach for air quality prediction in Morocco: combining CatBoost with metaheuristic optimization algorithmsFlood Prediction System: A Validation-First Machine Learning Pipeline for Flood Risk Assessment in Data-Scarce Environments

Short Term Power Demand Prediction Using Stochastic Gradient Boosting

Power prediction demand is vital in power system and delivery engineering fields. By efficiently pre

Hybridized light gradient boosting and whale optimization algorithm for diabetes detection

Due to their remarkable precision and effectiveness, gradient-boosted tree models have become the go

Public Debt Forecasting in Kenya: Integrating Macroeconomic Variables with Arimax and Extreme Gradient Boosting

In recent years, public debt in Kenya has been increasing rapidly, raising concerns about its sustai

Crop yield prediction in Ethiopia using gradient boosting regression

Nowadays, machine learning algorithms and methods are used in multiple areas of studies to achieve p

A hybrid Machine Learning approach for air quality prediction in Morocco: combining CatBoost with metaheuristic optimization algorithms

Air pollution poses serious risks to public health and environmental sustainability, particularly in

Flood Prediction System: A Validation-First Machine Learning Pipeline for Flood Risk Assessment in Data-Scarce Environments

Abstract Flooding remains one of the most destructive natural hazards in Cameroon, causing signific