ML-Based Soil Analysis System (Capstone / project work)
Other
Developed application-layer AI-enabled functionality as part of an ML soil-analysis system project under a broader capstone portfolio. Prepared and deployed code components for data-driven prediction using established ML algorithms. Ensured the solution met evaluation targets through model performance reporting. • Built prediction workflow using Python-based ML tooling. • Leveraged Random Forest, SVM, and Neural Networks for NIR-based soil property prediction. • Reported model performance metrics including R² values for pH, potassium, and phosphorus. • Included PCA-based variance retention (92.4% with 3 components) to support preprocessing.
2024 - 2025