💧 Water Potability Analyzer

---

🚀 Overview

Water Potability Analyzer is a machine learning-powered web application that predicts whether water is safe for consumption based on 9 key water quality parameters.

The system uses a Random Forest Classifier along with data preprocessing and feature analysis to provide accurate predictions and actionable insights.

👉 Designed for real-world applications like:

• Water quality monitoring
• Environmental analysis
• Public health assessment

---

🧠 Model Architecture

Component	Details
Model	Random Forest Classifier
Features	9 Water Quality Parameters
Preprocessing	Missing value imputation + StandardScaler
Optimization	GridSearchCV
Evaluation	Cross-validation + Accuracy Score

---

🧪 Water Quality Parameters

• pH
• Hardness
• Total Solids
• Chloramines
• Sulfate
• Conductivity
• Organic Carbon
• Trihalomethanes
• Turbidity

---

📊 Features of the Application

✔️ Manual parameter input for prediction
✔️ Natural language-based water analysis
✔️ Batch prediction using CSV upload
✔️ Feature importance & RFE analysis
✔️ Data visualization & insights dashboard
✔️ Water Quality Index (WQI) categorization
✔️ Interactive charts using Plotly

---

📊 Dataset

• 📊 Real-world water quality dataset
• 🧪 9 chemical parameters
• 🏷️ Target: Potability (Safe / Unsafe)
• 📌 Missing values handled using mean imputation

---

⚙️ Tech Stack

Layer	Technology
Frontend	Streamlit
Backend	Python
ML Libraries	Scikit-learn
Data Processing	Pandas, NumPy
Visualization	Plotly, Matplotlib, Seaborn
Model Persistence	Pickle

---

📱 Application Preview

🏠 Manual Input & Prediction

---

💬 Natural Language Analysis

---

📁 Batch Analysis (CSV Upload)

---

📊 Results & Recommendations

---

🔄 Working Flow

1. 📝 User inputs parameters / description / CSV
2. ⚙️ Data preprocessing (scaling + cleaning)
3. 🌲 Random Forest model predicts potability
4. 📊 Probability and confidence score generated
5. 📈 Results visualized using interactive charts
6. 💡 Recommendations provided based on prediction

---

📈 Model Training Highlights

• Compared multiple models (Logistic Regression, SVM, KNN, Decision Tree, Random Forest, etc.)
• Selected Random Forest based on best performance
• Applied GridSearchCV for hyperparameter tuning
• Used Stratified Train-Test Split for balanced evaluation
• Achieved strong classification performance

📌 Training & model pipeline implemented → :contentReference[oaicite:0]{index=0}

---

⚙️ Run Locally

git clone https://github.com/puniaiml/Water-potability-analysis.git
cd Water-potability-analysis
pip install -r requirements.txt
streamlit run app.py