Neural Information Processing Systems· 2016
Supervised Learning with Tensor Networks
- 214citations
- 2016year
Short summary
Tensor networks, specifically matrix product states (tensor trains), can parameterize non-linear kernel learning models, achieving <1% test error on the MNIST dataset.
AI-generated from the title and abstract; the full text is not read.
Key points
- Tensor networks, specifically matrix product states (tensor trains), are used to parameterize non-linear kernel learning models.
- The method achieved a test set classification error of less than 1% on the MNIST dataset.
- The tensor network structure imparts additional, interpretable structure to the learned model.
AI-generated from the title and abstract; the full text is not read.
Abstract
Tensor networks are approximations of high-order tensors which are efficient to work with and have been very successful for physics and mathematics applications. We demonstrate how algorithms for optimizing tensor networks can be adapted to supervised learning tasks by using matrix product states (tensor trains) to parameterize non-linear kernel learning models. For the MNIST data set we obtain less than 1% test set classification error. We discuss an interpretation of the additional structure imparted by the tensor network to the learned model.
The authors' abstract, as published at the source. Neural Information Processing Systems, 2016 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
Field: Computational Mathematics
Computational MathematicsMathematics