Movie Likeness Score

I developed advanced neural network-based machine learning models to predict new ratings from user-generated similarity scores, focusing on enhancing recommendation accuracy. This involved designing and training neural networks to interpret and learn from large datasets of user reviews and similarity scores. Using the Rotten Tomatoes dataset, which comprised 1,129,887 reviews, I meticulously preprocessed the data to ensure quality and relevance for model training.

The program I created achieved an impressive 93.3% accuracy in predicting user ratings, significantly improving the reliability of recommendations. This high level of accuracy was achieved by carefully tuning the neural network architecture and hyperparameters to optimize performance. The system provides users with personalized and precise rating predictions, enhancing their overall experience by recommending content that aligns closely with their preferences.