Deep Transfer Tensor Factorization for Multi-View Learning

TL;DR

This paper introduces Deep Transfer Tensor Factorization (DTTF), a novel approach combining deep learning and tensor factorization to address data sparsity in multi-view learning, significantly improving rating prediction accuracy.

Contribution

The paper proposes a generic deep transfer tensor factorization architecture that integrates deep learning with cross-domain tensor factorization, effectively utilizing side information to mitigate data sparsity.

Findings

DTTF outperforms state-of-the-art methods on real-world datasets.

The integration of deep autoencoders with tensor factorization improves recommendation accuracy.

Side information enhances the model's ability to handle sparse multi-view data.

Abstract

This paper studies the data sparsity problem in multi-view learning. To solve data sparsity problem in multiview ratings, we propose a generic architecture of deep transfer tensor factorization (DTTF) by integrating deep learning and cross-domain tensor factorization, where the side information is embedded to provide effective compensation for the tensor sparsity. Then we exhibit instantiation of our architecture by combining stacked denoising autoencoder (SDAE) and CANDECOMP/ PARAFAC (CP) tensor factorization in both source and target domains, where the side information of both users and items is tightly coupled with the sparse multi-view ratings and the latent factors are learned based on the joint optimization. We tightly couple the multi-view ratings and the side information to improve cross-domain tensor factorization based recommendations. Experimental results on real-world datasets demonstrate that our DTTF schemes outperform state-of-the-art methods on multi-view rating predictions.