This project is a movie recommender system powered by deep learning. I used content-based filtering with movies being scored out of 5 for their relevance across various genres and users ranking their interest in those same genres. The model has been trained on a dataset from MovieLens, and the neural networks are built using TensorFlow.
- Content-based filtering system powered by TensorFlow
- Chronological documentation from data cleaning to model testing
- Ability to obtain personal recommendations using the model
- TensorFlow/Keras
- NumPy
- Pandas
- Scikit-learn
- Jupyter Notebook
To try out the movie recommender, download the repository and open code.ipynb. Run all of the cells in 'Building the Model', and then scroll down to 'Predictions for a New User'. Input your genre preferences (scored out of 5) and run both cells.