How minakfahmy/ma_water_quality_analysis works
This repository is a Massachusetts water-quality analysis project, not a deployed monitoring service. Its documented workflow uses the probe-data CSV in an end-to-end notebook to clean and explore measurements, train risk classifiers, tune the XGBoost decision threshold, and evaluate high-risk-event predictions. The notebook also documents class imbalance and predictor findings. No source excerpts were available, so the graph reflects README-described relationships only; it does not imply that predictions are delivered to agencies or that a live alerting system exists.
Main entry points: MA_Water_Quality_Report.ipynb Analysis notebook · wqdiscreteprobedata.csv Probe data
Components 10
Water-quality data
- Probe data
wqdiscreteprobedata.csv
Notebook analysis
- Analysis notebook
MA_Water_Quality_Report.ipynb - Data cleaning
MA_Water_Quality_Report.ipynb - Exploratory analysis
MA_Water_Quality_Report.ipynb
Risk modeling
- Feature preparation
MA_Water_Quality_Report.ipynb - Logistic baseline
MA_Water_Quality_Report.ipynb - XGBoost model
MA_Water_Quality_Report.ipynb - Threshold tuning
MA_Water_Quality_Report.ipynb - Model evaluation
MA_Water_Quality_Report.ipynb
Other components
- Analyst
Connections 9
- Analyst → Analysis notebook: runs
- Probe data → Analysis notebook: provides measurements
- Analysis notebook → Data cleaning: performs
- Analysis notebook → Exploratory analysis: performs
- Analysis notebook → Feature preparation: prepares features
- Analysis notebook → Logistic baseline: trains baseline
- Analysis notebook → XGBoost model: trains model
- Analysis notebook → Threshold tuning: tunes threshold
- Analysis notebook → Model evaluation: evaluates models