Registration Open

Data Science Hackathon — Air Quality Prediction

BUILD • INNOVATE • COMPETE • WIN

Teams of 1–6

Develop a data science solution that analyzes historical environmental data and predicts future Air Quality Index (AQI).

Registration closes in

--Days

--Hours

--Minutes

--Seconds

1
Registration
24 Aug, 5:34 pm – 29 Aug, 4:30 am
2
Submissions
29 Aug, 4:30 am – 30 Aug, 4:30 pm
3
Results
Within 2 day(s) of close

Problem Statement

Air pollution changes based on factors such as weather, traffic, industrial activity, and historical pollution patterns. Predicting future air quality can help people and authorities take preventive action.

What Students Need to Create

Develop a data science solution that analyzes historical environmental data and predicts future Air Quality Index (AQI).

Students should implement:

Data Collection

Use datasets containing:

AQI
PM2.5
PM10
CO
NO₂
SO₂
O₃
Temperature
Humidity
Wind speed

Data Cleaning

Handle missing values.
Remove duplicates.
Detect outliers.
Prepare the dataset for analysis.

Exploratory Data Analysis

Analyze:

Pollution trends
Seasonal patterns
Weather vs pollution
Pollutant relationships
High-pollution periods

Prediction Model

Build a machine-learning model to predict:

Future AQI
Pollution levels
AQI category (example categories: Good → Moderate → Unhealthy → Very Unhealthy → Hazardous)

Model Evaluation

Demonstrate appropriate evaluation metrics such as:

MAE
RMSE
R²
Accuracy/F1 if using classification

Visualization

Create visualizations for:

Historical AQI
Pollution trends
Predicted AQI
Pollutant contribution
Location-wise pollution

Interactive Dashboard

Allow users to:

Select a location.
Select a date/time period.
View historical AQI.
View predicted AQI.
Understand major pollution factors.

Expected Output

A data-driven application that demonstrates: Data Collection → Data Cleaning → EDA → Feature Engineering → ML Model → Prediction → Visualization → Insights

What to Submit

Submit all of the following:

GitHub repository link with the complete project source code.
Full documentation of the project (PDF/DOC) with a clear explanation of the outcomes, including screenshots of the output. If the team has more than one member, every member's name must be listed in the documentation.
No live link is required for this track — the documentation and screenshots are sufficient.

Rules

Any free public air-quality/weather dataset may be used (e.g. government open-data portals, Kaggle datasets).

Prizes

🥇

1st Prize

₹3,000

InternshipWinner Certificate

🥈

2nd Prize

₹2,000

InternshipWinner Certificate

🥉

3rd Prize

₹1,000

Certificate