Predicting IMDb Ratings with Linear Regression

Contents

• Notebook to run through my modeling process
  
• Folder of .py files to properly run the notebook
  
• Exact data I used for to create the model
  
• Visualizations used in my final presentation
  
• Slides for my final presentation of the model
  

Extras: Read the blog post

Description

This repository contains a working model to predict IMDb ratings for a movie using features available prior to the movie's release. The model uses linear regression and features obtained through scraping movie information from IMDb using BeautifulSoup.

Features and Target Variables

• Target Variable: IMDb Rating
• Features: Runtime, MPAA Rating, Genre, Director, Writer, Stars, Production Company, Release Month, Years Since Release

Data Used

Scraped over 8,000 IMDb pages to collect movie data.

Tools Used

• Beautiful Soup for web scraping
• Linear regression
• Ridge regression
• Polynomial regression
• Supervised Machine Learning
• Feature Engineering & Selection
• Numpy
• Pandas
• Seaborn
• Matplotlib

Potential Impact

This model is a good basis for producers or movie enthusiasts to understand what rating a movie will get after it is released based on variables that can all be determined prior to the release of the movie. Below is a visual illustrating the results from the analysis, highlighting what features had the largest effect on the model.