Predicting OTT Subscription Behavior using Deep Learning: A Multi-Class Classification Approach with Feature Engineering and Class Imbalance Handling

Key Points

• The multi-class classification model predicts subscription behavior effectively, increasing understanding of consumer preferences.
• Important factors of feature engineering include content genres, pricing models, and user demographics in the analysis.
• This observational analysis employed deep learning techniques to address class imbalance and enhance prediction accuracy.
• Findings support more tailored marketing strategies and highlight the importance of data-driven decision-making in media.