BTRS → BNGRS coordinate transformation suite using XGBoost + Neural Network ensemble. High-accuracy geodetic conversion for Botswana reference systems with web interface, batch processing, and interactive mapping.
# Bayesian Optimization ANN for Geodetic Coordinate Transformation
This project implements an Artificial Neural Network (ANN) optimized using Bayesian Optimization techniques for high-precision coordinate transformation between Botswana geodetic reference systems.
## Features
- ANN regression model
- Bayesian hyperparameter optimization
- Coordinate transformation workflow
- Python implementation using TensorFlow/Keras
- Model evaluation and accuracy analysis
## Technologies Used
- Python
- TensorFlow / Keras
- Scikit-learn
- NumPy
- Pandas
- Bayesian Optimization
## Research Context
Developed as part of geospatial/geodetic research focused on transforming coordinates between legacy and modern reference systems in Botswana.
## Author
Theo Madziba