Linear Regression Models in Python

The Jupyter Notebook is available here to download.

Introduction • Summary • Conclusion & Plans

Introduction

This notebook will go over how to apply models to predict purchase amount using the BlackFriday dataset off Kaggle. You can access this dataset by clicking on this link. This project will be done in 6 steps shown below.

• Data Collection
• Processing Data
• Data Cleansing
• EDA
• Model building
• Communication & Visualization

Summary

After going through the different steps in the project, we ended up building several different regression models and used both forward and backward elimination to come up with an optimal models. Variable selection methods such as forward elimination do not have any native support from any of the Python libraries, so this was built in.

Conclusion & Plans

In the future, I will continue to add new methods for the model building phase especially for model selection. I will also go more into depth on how to collect this type of data, since the data collection process can be complex.