Explainable Artificial Intelligence integrated with Machine learning operations to predict the nitrate concentrations in Groundwater DOI Creative Commons
Jagadish Kumar Mogaraju

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 27, 2025

Abstract Groundwater is a commodity we depend on for diverse needs, and maintaining its quality must be considered vital. We Machine Learning (ML) operations Explainable Artificial Intelligence (XAI) to predict the nitrate concentration levels in groundwater of India years 2019 2023. The variables used this study are Latitude, Longitude, pH, EC, CO3, HCO3, Cl, SO4, PO4, TH, Ca, Mg, Na, K, F, TDS, SiO2, NO3 dataset Fe, As, U, 2023 dataset. prepared GIS surface maps using interpolation supported by Empirical Bayesian Kriging method. investigated model efficiency feature importance presence absence location attributes. 19 ML models filtered Light Gradient Boosting (LightGBM) Liner Regression (LR) that exhibited relatively better accuracy. first trained these fed them XAI via SHAP (SHapley Additive exPlanations), which was dependent game theory. obtained 28.23% 24.88% increase accuracy when comparing datasets with attributes, respectively. also observed 28.3% without attribute used. conclude can integrated improve prediction studies.

Language: Английский

Explainable Artificial Intelligence integrated with Machine learning operations to predict the nitrate concentrations in Groundwater DOI Creative Commons
Jagadish Kumar Mogaraju

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 27, 2025

Abstract Groundwater is a commodity we depend on for diverse needs, and maintaining its quality must be considered vital. We Machine Learning (ML) operations Explainable Artificial Intelligence (XAI) to predict the nitrate concentration levels in groundwater of India years 2019 2023. The variables used this study are Latitude, Longitude, pH, EC, CO3, HCO3, Cl, SO4, PO4, TH, Ca, Mg, Na, K, F, TDS, SiO2, NO3 dataset Fe, As, U, 2023 dataset. prepared GIS surface maps using interpolation supported by Empirical Bayesian Kriging method. investigated model efficiency feature importance presence absence location attributes. 19 ML models filtered Light Gradient Boosting (LightGBM) Liner Regression (LR) that exhibited relatively better accuracy. first trained these fed them XAI via SHAP (SHapley Additive exPlanations), which was dependent game theory. obtained 28.23% 24.88% increase accuracy when comparing datasets with attributes, respectively. also observed 28.3% without attribute used. conclude can integrated improve prediction studies.

Language: Английский

Citations

0