Ever wondered why Hollywood, despite employing 13,500+ actors and netting a $150 billion profit this year, can still be capable of producing mediocre movies? The reason behind this is because Hollywood relies on marketing to make movies successful. Marketing doesn’t always work, though, with each mediocre film costing hundreds of millions in lost revenue. What if there was a way to predict how well a film would fare before it was released to eliminate this risk? After our team searched through possible projects that could maximize commercial profits, it was decided to take on predicting movie popularity! A machine learning model can be built, which is a program that looks at old data to find a relationship, then uses it on new data. In this case, we will
program a model to look at old movie data to determine the relationship between those movies’ information (director, actors, budget, etc) and how well they did on their IMDb ratings. It will then use this discovered relationship to predict how successful new movies will be using their IMDb based on their data as well. This way, Hollywood could potentially eliminate producing mediocre movies and focus solely on the ones predicted to be good ones. In the end, our model accurately predicted the IMDb scores of recently published movies as well as new ones.