Air Quality prediction by using Machine learning

Predicting Air quality by using Machine learning supervised algorithms.

VijayAnaparthi/Air-Quality-Prediction

About Data set(AirPi Data — AirPi.csv file)

1. Data set contains 9 features and 1 target variable i.e air_quality.
2. It contains 14572 data points and i divide data into 70%(train) and 30%(test).
3. Basically i am not did any exploratory data analysis on this data set. If you are interested you can contribute to that.

Trained models

It is a regression problem so i trained 3 models i.e

1. Linear regression
2. Random forest regressor
3. XGboost regressor

In this process i also did hyper parameter tuning by using Randomsearchcv then trained the model again.

Performance metrices

1. Mean squared error(MSE)
2. Mean absolute percentage error(MAPE)
3. R2 score

Conclusion

1. I found out XGboost regressor is less overfitting compare to remaining two models. But Random forest regressor giving the best results in terms of error.
2. So in above Airpi_output.csv file i added 3 more columns and appended predicted values of every model for each data point.