Water depletion predictor

This repo is a groundwater quantity prediction ML app with a small web frontend and a Flask-style backend wrapped around a trained model.[github]​

Core idea

• Predicts water quantity from physicochemical features using a trained ML model saved as model.pkl.

• Uses a CSV dataset water_data.csv and Jupyter notebooks to explore data and train the model.

Repo structure

• Root:

  • README.md: minimal placeholder (“water-prediction-ml”).

  • GroundWater_Project/: actual project code and assets.

• GroundWater_Project/ contents:

  • GroundWater_DataSet.ipynb: data loading, cleaning, and exploration notebook.

  • GroundWater_TrainedModel.ipynb: model training and evaluation notebook.

  • .ipynb_checkpoints/: auto-saved notebook checkpoints.

  • water_data.csv: groundwater dataset used for training.

  • model.pkl: serialized trained model.

  • app.py: backend server exposing prediction endpoint(s).

  • index.html, index.js, main.css: simple web UI that calls the backend and displays predictions.

Backend behavior (inferred from files)

• app.py likely:

  • Loads model.pkl at startup.

  • Exposes an HTTP endpoint (probably /predict) to accept feature inputs (e.g., via JSON or form).

  • Runs the model to return predicted water quality or class.

Frontend behavior

• index.html: main page with input fields for water parameters and a section to show result.

• index.js: sends user inputs to the backend endpoint, handles the response, and updates the DOM.

• main.css: basic styling for the groundwater prediction UI.

Tech stack and ownership

• Primary language: Jupyter Notebook (98.2%) for ML experimentation, plus Python, HTML, CSS, JS for deployment.[github]​

• Contributors: edlynjessica and jebasinghsunderson (you), with recent work “Almost done with front end and Backend”.[github]​