⬅ Back to Portfolio

Movie Revenue & Rating Prediction

Machine Learning System  ·  Solo Project  ·  Python / Scikit-learn

A supervised machine learning system that predicts box office revenue and identifies optimal runtime ranges for maximizing audience ratings across film genres.

Overview

This system analyzes a dataset of ~4,800 film records to deliver two capabilities: revenue prediction based on budget, genre, and runtime, and runtime optimization analysis to identify the runtime ranges most correlated with high audience ratings per genre. Feature engineering and model selection were tuned to balance accuracy with interpretability.

Tech Stack

Python  ·  Pandas  ·  NumPy  ·  Scikit-learn  ·  Random Forest  ·  Gradient Boosting  ·  Jupyter Notebook  ·  Git

Model Performance

Runtime–Rating Correlations by Genre

Genre	Correlation (r)	Strength
Action	0.468	Strong
Thriller	0.459	Strong
Adventure	Moderate	Moderate
Drama	Moderate	Moderate
Horror	Moderate	Moderate
Comedy	Weak	Weak
Animation	Weak	Weak