Multi-expert neural networks for recommendation
Takeshi Kawato · Digital WPI · 2003
This project investigated a mixture of experts neural architecture for a combined collaborative and content-based recommender system. The effect of first reducing the dimensionality of the input data using the singular value decomposition was also studied. We showed that the mixture of experts architecture achieves the same recommendation quality as a fully-connected architecture while requiring less computation time, or, if desired, higher quality can be achieved with only slight increase in running time.