This is an entry for a Kaggle competition that was created as part as part of the AWS Machine Learning Engineer nanodegree.
This project sought to build an algorithm to enter in the Kaggle Bike Sharing Demand competition. A full project overview can be found at the link above. The competition uses data from Capital Bikeshare to forecast the usage of the city bikeshare system.
Using the AutoGluon framework I built a baseline model, then used feature engineering and hyperparameter tuning to improve upon the initial model.
The documents contained in this repository are below:
Analysis - this file contains the actual code that was ran to generate the machine learning models
Project Report - this file contains a write up with some questions answered regarding the analysis