Are you a lover of 20th century movies? If yes, then this tool is for you. This tool will recommend movies on the basis of genre, director, actors etc. Autoencoders is the technique used here, it will learn correlations by recreating ratings of every customer. Then it will improve its predictions by backpropagating the errors.
I have written an article on recommendation system using autoencoders here
This article will help you understand all the concepts including maths, network engineering etc of autoencoders.
- Clone this repo to your computer.
- Get into the folder using cd Movie-Recommender-System.
- Download the data files from MovieLens into this directory.
In your terminal type
- pip3 install torch torchvision to install with pip
- conda install pytorch torchvision -c pytorch to install with Anaconda
Run each cell in the Recommendation System.ipynb file