Locally connected deep learning framework for industrial-scale recommender systems
Proceedings of the 26th international conference on World Wide Web companion, 2017•dl.acm.org
In this work, we propose a locally connected deep learning framework for recommender
systems, which reduces the complexity of deep neural network (DNN) by two to three orders
of magnitude. We further extend the framework using the idea of the recently proposed
Wide&Deep model. Experiments on industrial-scale datasets show that our methods could
achieve good results with much shorter runtime.
systems, which reduces the complexity of deep neural network (DNN) by two to three orders
of magnitude. We further extend the framework using the idea of the recently proposed
Wide&Deep model. Experiments on industrial-scale datasets show that our methods could
achieve good results with much shorter runtime.
In this work, we propose a locally connected deep learning framework for recommender systems, which reduces the complexity of deep neural network (DNN) by two to three orders of magnitude. We further extend the framework using the idea of the recently proposed Wide&Deep model. Experiments on industrial-scale datasets show that our methods could achieve good results with much shorter runtime.
ACM Digital Library