A Recommender system is a useful engine to predict things according to our interests. This paper explains the recommendation system which is based on the State of the Art Deep AutoEncoders which comes under Model-based filtering technique. Deep AutoEncoders are Artificial Neural Network that is used in Computer Vision and NLP. Here Deep AutoEncoders are used to find Top N recommendation of movies by an Unsupervised learning method. Since the patterns are found using Collaborative Filtering, Netflix three months dataset with implicit values have been used for evaluating the proposed recommender system. This paper is good at predicting the ratings missing from the sparse matrices which is initially provided as input. 20 models have been designed and the best model is linked with the frontend. The RMSE value for the training set and test set is 0.8686 and 0.9365 respectively which tells that this model does not overfit. It is thus proved that Deep AutoEncoders generalizes well.
Movie Recommendation System Using Deep Autoencoder
2021-12-02
976133 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
VISUALIZATION OF DRIVING BEHAVIOR USING DEEP SPARSE AUTOENCODER
British Library Conference Proceedings | 2014
|Asphalt Concrete Mix Design Optimization Using Autoencoder Deep Neural Networks
Transportation Research Record | 2023
|Rail Defect Detection Using Ultrasonic A-Scan Data and Deep Autoencoder
Transportation Research Record | 2023
|