ML project
This personal machine learning project was undertaken as part of the HarvardX Data Science Professional Certificate program on edx.org.
The dataset was found on kaggle and uses publicly available bike ride counter data from the City of Ottawa (Canada) married with weather data from Environment Canada.
The weather and other features, such as day of the year, were used to make predictions about the number of bike riders per day passing by various counters in the Ottawa area.
The data can be found in zip format on Kaggle at the following address: https://www.kaggle.com/m7homson/ottawa-bike-counters/download
To access it on Kaggle you have to create an account and login.
The datasets have also been saved to this repo as have the data import and wrangling script, the final script for the project and the project report in RMD and PDF formats.
Environment and Climate Change Canada weather data is subject to Government of Canada Open Government Data Licence 2.0.
Ottawa City Bike Counter data is subject to City of Ottawa Open Data Licence 2.0.
Update July 2021 Added an update script which includes weather and bike counter data for 2019 to further evaluate and explore the model.
The updated bike counter data can be found at the City of Ottawa open data portal: https://open.ottawa.ca/documents/ottawa::bicycle-trip-counters/about
Data is in an MSExcel spreadsheet format. A copy of the spreadsheet used for this analysis is uploaded to the repo.