FDM_Mini_Project
Overview
Welcome to the Water Quality Prediction System, a tool that predicts the potability of water samples based on various parameters. Whether you are a policymaker, environmental agency, water treatment plant, researcher, or part of a community, this system can help you assess the quality of water samples you've collected.
Technologies Used
- Frontend: Developed with React.js and Vite for a smooth user experience.
- Backend: Powered by Azure Serverless Function Apps, ensuring scalability and efficiency.
- Machine Learning Models: Built using Python's scikit-learn library to provide accurate predictions.
Model Training
- Six models were trained: Logistic Regression, Support Vector Machine (SVM), K-Nearest Neighbors (K-NN), Decision Tree, Random Forest, and XGBoost.
- After thorough evaluation, the Random Forest model was selected for its superior performance.
- Data preprocessing included handling missing values, categorical variables, normalization of numeric values, and addressing outliers.
How to Use
- Access the web application here.
- Input water quality parameters, such as pH, iron content, and turbidity.
- Receive a prediction on the potability of your water sample.
Installation
To set up the project locally, follow these steps:
- Clone the repository:
- Install dependencies for the frontend and backend (npm install)
- Run the application (FE: npm run dev)
Acknowledgments
- Water quality dataset source: Kaggle