My project applied data analysis to explore what factors influence movie ratings and generate insights to guide better decision-making in the film industry.
Predictive Insights into Movie Ratings
I worked on a group project analyzing 700k+ movies with 20 features to uncover insights into the movie industry and predict audience ratings.
The goal was to guide production decisions such as budgeting, release scheduling, and content strategy to maximize engagement and financial sustainability.
Project Overview
Analyzed 700k+ movies with 20 features to predict audience ratings and guide production decisions (budgeting, release timing, content strategy).
Explored trends in genres, budgets, and audience engagement to uncover insights driving movie success.
Cleaned data: split release dates, removed irrelevant columns, filtered nulls and zeros.
Created visualizations to examine genre popularity, potential biases, and correlations.
Modeling
After building and comparing predictive models such as linear regression (baseline), decision tree regressor, KNN, and random forest, we selected decision tree regressor due to its lower RMSE, reduced overfitting, and better interpretability.
Case Study
To put our predictive model to test, we created a case study including ratings for two hypothetical movies. We recommended producing the movie with the higher predicted vote_average. We then compared our predicted ratings to genre averages to validate our models performance.
Key Learnings & Takeaways
I learned about data cleaning, EDA, model evaluation, and applying predictive analytics to real-world business decisions. It was interesting to see how predictive insights help optimize creative and financial outcomes, and model interpretability is critical for actionable recommendations.