This project was to perform analysis for Adventure Works Bikes customer data.
It was posed during Microsoft's Principles of Machine Learning Certification and I thought it was interesting.
From the available data provided - Cutomer Information, Customer Spending and Customer Bike Buyer data. We needed to provide two things:
- We need to predict if a customer will buy a bike or not given the above information.
- We need to predict customer average month spend.
I explore the bike buy or not problem in the "Classification" notebook and the average month spend in the "Regression" notebook.
Various algorithms and ensemble methods were of interest for this particular data such as: RandomForest classifier and regressor, Logistic Regression, NeuralNetworkRegressor and Ploynomial Regression. These were done in conjunction with Cross Validation for hyperparameter tuning and creative feature generation.
Feel free to take a look and let me know if you make any additional discoveries or have comments.