🌫️ Air Quality Index (AQI) Prediction using Python

Predicting Air Quality Index (AQI) using machine learning techniques in Python — analyzing pollutant data to forecast air quality levels and help identify environmental health risks.

---

📌 Table of Contents

• Overview
• What is AQI?
• Dataset
• Tech Stack
• Project Structure
• Workflow
• Models Used
• Results
• How to Run
• Author

---

📖 Overview

Air pollution is one of the most critical environmental challenges of the 21st century. This project builds a machine learning model to predict the Air Quality Index (AQI) based on concentrations of key air pollutants. By accurately predicting AQI, we can:

• Provide early warnings for hazardous air quality events
• Help governments and citizens make informed decisions
• Analyze pollution trends across regions and time periods

---

🌍 What is AQI?

The Air Quality Index (AQI) is a standardized scale used to communicate how polluted the air currently is or how polluted it is forecast to become.

AQI Range	Category	Health Impact
0 – 50	Good	Little or no risk
51 – 100	Moderate	Acceptable; some concern for sensitive groups
101 – 150	Unhealthy for Sensitive Groups	Sensitive people may experience effects
151 – 200	Unhealthy	Everyone may begin to experience effects
201 – 300	Very Unhealthy	Health alert — serious effects for everyone
301+	Hazardous	Emergency conditions

---

📊 Dataset

The dataset contains readings of major air pollutants used to compute AQI:

Feature	Description
PM2.5	Fine particulate matter (≤ 2.5 µm)
PM10	Coarse particulate matter (≤ 10 µm)
NO	Nitric Oxide
NO2	Nitrogen Dioxide
NOx	Nitrogen Oxides
NH3	Ammonia
CO	Carbon Monoxide
SO2	Sulfur Dioxide
O3	Ozone
Benzene	Benzene concentration
Toluene	Toluene concentration
AQI	Target variable — Air Quality Index
AQI_Bucket	AQI category label

---

🛠️ Tech Stack

Python 3.8+
├── pandas          — Data loading & manipulation
├── numpy           — Numerical operations
├── matplotlib      — Data visualization
├── seaborn         — Statistical plots
├── scikit-learn    — ML models & preprocessing
└── Jupyter Notebook — Interactive development environment

---

📁 Project Structure

aqi-prediction-python/
│
├── Predicting_Air_Quality_Index_using_Python.ipynb   # Main notebook
└── README.md                                          # Project documentation

---

🔄 Workflow

1. Data Loading
        ↓
2. Exploratory Data Analysis (EDA)
   ├── Shape, dtypes, null values
   ├── Distribution plots
   └── Correlation heatmap
        ↓
3. Data Preprocessing
   ├── Handling missing values
   ├── Feature selection
   └── Train-test split
        ↓
4. Model Training
   ├── Multiple regression models
   └── Hyperparameter tuning
        ↓
5. Model Evaluation
   ├── R² Score
   ├── MAE / RMSE
   └── Prediction vs Actual plots
        ↓
6. Results & Conclusions

---

🤖 Models Used

Model	Type
Linear Regression	Baseline regression
Random Forest Regressor	Ensemble — tree based
Decision Tree Regressor	Tree based
K-Nearest Neighbors	Instance based

---

📈 Results

The models were evaluated using standard regression metrics:

• R² Score — measures how well predictions fit actual AQI values
• MAE (Mean Absolute Error) — average prediction error magnitude
• RMSE (Root Mean Squared Error) — penalizes larger errors more heavily

Refer to the notebook for detailed metric comparisons and visualization plots.

---

▶️ How to Run

1. Clone the Repository

git clone https://github.com/uddhav05-cyber/aqi-prediction-python.git
cd aqi-prediction-python

2. Install Dependencies

pip install pandas numpy matplotlib seaborn scikit-learn jupyter

3. Launch Jupyter Notebook

jupyter notebook Predicting_Air_Quality_Index_using_Python.ipynb

4. Run All Cells

In Jupyter: Kernel → Restart & Run All

---

👨‍💻 Author

Uddhav Bhople

BTech Computer Engineering | DY Patil University, Pune Software Engineering Student & Aspiring AI Engineer

---

📄 License

This project is licensed under the MIT License.

---

If you found this project helpful, please consider giving it a ⭐