Predicting Box Office Success with Advanced Regression Techniques: Using Data to Merge Art and Analytics

Kawade, Shubham Pandurang (2025) Predicting Box Office Success with Advanced Regression Techniques: Using Data to Merge Art and Analytics. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (776kB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

The film industry seeks data-driven strategies to predict box office success, motivating this thesis to investigate key predictors of revenue and IMDb ratings using the Kaggle IMDb Movie Dataset (1,000 films). The research question asks: What are the key predictors of box office revenue and ratings, and how effectively can regression and tree-based models predict these outcomes? Objectives include identifying predictors, evaluating models, and offering industry insights. Using Python (scikit-learn, xgboost), the study pre-processed data (imputation, genre encoding), engineered features (e.g., rating_votes_interaction), and applied Linear Regression, Ridge, Lasso, Random Forest, Gradient Boosting, and XGBoost. Core findings show Linear Regression predicts revenue effectively (Test R² ~63%, within 50–70% benchmarks), driven by audience engagement (Votes, 26.74% importance) and genres (Animation: $184.1M, Adventure). Rating prediction failed (Test R² ~9%, below 20–30% benchmarks), lacking features for subjective outcomes. Academically, the study highlights linear models’ efficacy and feature engineering value, while practitioners can prioritize Animation and marketing to boost engagement. Limitations include missing budget data and non-random sampling. Future work proposes integrating budgets, sentiment analysis from X posts, and larger datasets. This framework offers potential for a commercial forecasting tool, enhancing film industry decision-making.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Zahoor, Sheresh UNSPECIFIED
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Film Industry Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	20 Aug 2026 09:43
Last Modified:	20 Aug 2026 09:43
URI:	https://norma.ncirl.ie/id/eprint/9561

Actions (login required)

View Item