A Neural Autoregressive Approach to Collaborative Filtering

Abstract

This paper proposes CF-NADE, a neural autoregressive architecture for collaborative filtering (CF) tasks, which is inspired by the Restricted Boltzmann Machine (RBM) based CF model and the Neural Autoregressive Distribution Estimator (NADE). We first describe the basic CF-NADE model for CF tasks. Then we propose to improve the model by sharing parameters between different ratings. A factored version of CF-NADE is also proposed for better scalability. Furthermore, we take the ordinal nature of the preferences into consideration and propose an ordinal cost to optimize CF-NADE, which shows superior performance. Finally, CF-NADE can be extended to a deep model, with only moderately increased computational complexity. Experimental results show that CF-NADE with a single hidden layer beats all previous state-of-the-art methods on MovieLens 1M, MovieLens 10M, and Netflix datasets, and adding more hidden layers can further improve the performance.

Accepted by ICML2016

References in corpus (7)

• Adam: A Method for Stochastic Optimization
• Going Deeper with Convolutions
• Neural Network Matrix Factorization
• Blocks and Fuel: Frameworks for deep learning
• Scalable Recommendation with Poisson Factorization
• A Deep and Tractable Density Estimator
• A Deep and Autoregressive Approach for Topic Modeling of Multimodal Data

Cited by in corpus (31)

• Deep Learning based Recommender System: A Survey and New Perspectives
• Graph Convolutional Matrix Completion
• Embarrassingly Shallow Autoencoders for Sparse Data
• Multi-Behavior Enhanced Recommendation with Cross-Interaction Collaborative Relation Modeling
• Multi-Behavior Graph Neural Networks for Recommender System
• Variational Autoencoders for Collaborative Filtering
• STAR-GCN: Stacked and Reconstructed Graph Convolutional Networks for Recommender Systems
• GLocal-K: Global and Local Kernels for Recommender Systems
• Inductive Matrix Completion Based on Graph Neural Networks
• Machine-learning-assisted Monte Carlo fails at sampling computationally hard problems
• Contemporary Recommendation Systems on Big Data and Their Applications: A Survey
• Recent Advances in Heterogeneous Relation Learning for Recommendation
• Joint Text Embedding for Personalized Content-based Recommendation
• DARec: Deep Domain Adaptation for Cross-Domain Recommendation via Transferring Rating Patterns
• Why Not to Use Zero Imputation? Correcting Sparsity Bias in Training Neural Networks
• Scalable Realistic Recommendation Datasets through Fractal Expansions
• Do Offline Metrics Predict Online Performance in Recommender Systems?
• Cross-domain Recommendation via Deep Domain Adaptation
• A Hybrid Variational Autoencoder for Collaborative Filtering
• Expert Finding in Community Question Answering: A Review
• Multi-Component Graph Convolutional Collaborative Filtering
• Spectral Geometric Matrix Completion
• On Sampling Strategies for Neural Network-based Collaborative Filtering
• JSCN: Joint Spectral Convolutional Network for Cross Domain Recommendation
• Towards Amortized Ranking-Critical Training for Collaborative Filtering
• Low-Rank Matrix Completion: A Contemporary Survey
• Multi-output Gaussian Processes for Uncertainty-aware Recommender Systems
• Scaling Up Collaborative Filtering Data Sets through Randomized Fractal Expansions
• MIC: Model-agnostic Integrated Cross-channel Recommenders
• Scalable Recommender Systems through Recursive Evidence Chains
• A Collective Variational Autoencoder for Top-$N$ Recommendation with Side Information