Air Quality Index Prediction

This project aims to predict AQI (PM 2.5) from several physical factors. The identified ML problem is the regression problem statement wherein we predict a numerical value. Several Models where used and the best result was achived with XGBoostRegressor+GridSearchCV. various models are used like Linear & Lasso Regression DecisionTree RegressorRandom forest Regrssor & XGBoost Regressor. An EDA is also peroformed on the data as well as feature importance.

Deployed on streamlit

https://air-quality-predict.streamlit.app/

Data

The data has been webscrapped from https://en.tutiempo.net/

Data is from 2013-2018 for most days in the month

Project Findings