The project focuses on predicting the amount of carbon dioxide emitted by vehicles using various Machine Learning algorithms.
The prediction will allow vehicle manufacturers have an idea about the amount of carbon dioxide emitted by the yet to be manufactured vehicles in future and manufacture them prudently.
The best performing model achieved an accuracy of 99.86% on the unseen data.
The dataset provides model-specific fuel consumption details and estimated carbon dioxide emissions for a large number of vehicles over many years in Canada.
The raw datasets were taken from the Fuel Consumption ratings webpage of the official Canada Government open data website.
The dataset in this repository was compiled from the raw datasets taken from the mentioned website. The data is spread across three different time periods namely 1995-1999, 2000-2014 and 2015-2020 in three different sheets of the Excel workbook.
A few libraries has to be installed in order to run the notebook in your local system.
Libraries required to run the notebook are mentioned in requirements.txt.
Note
You can also view the notebook on Google Colab where you can run the notebook and make changes if you want without downloading the notebook to your local system.
