This week, we wrapped up the linear regression topic by presenting projects in which we developed a model to predict a movie's success.
For my project, I looked at movies "Based on a True Story" and collected data to determine which features are the most influential to the Worldwide Gross of movies belonging to this category.
I used Beautiful Soup to scrape data from IMDB.com, BoxOfficeMojo.com, and TheNumbers.com, and found 10 features for movies Based on a True Story:
In order to develop an accurate predictive model, I began by performing Ordinary Least Squares regression incorporating all of my feautures in the model, and iteratively removing features that showed a p-value higher than a certain threshold (0.100). The initial models were applied to a training set of 70% of the movie data, and the results below show the models performance on the test set.
The final model showed that the most influential features were Total # Theaters, Total Gross, Budget, and the Romance genre:
This can be illustrated by the linear relationship between these feautures in the raw data as shown below:
The model predictions proved to be accurate overall, as can be seen by the Fit plot below (red = model prediction, blue = raw data, black = 95% confidence interval).
The residuals in this model increased in magnitude somewhat as the predicted Worldwide Gross increased, mainly because there was a lower density of data in this section for the model to predict. Additionally, it is always challenging to obtain accurate predictions for more extreme outcomes.
The results of the model imply the following conclusions for movies "Based on a True Story":
-Domestic release (measured by Total Theaters and Total Domestic Gross) is key indicator of Worldwide Gross
-Total Domestic Gross and Budget are linearly related to Worldwide Gross
-Romance genre is the most highly correlated to Worldwide Gross