Projects per year
Abstract
Tensors are increasingly encountered in prediction problems. We extend previous results for high-dimensional least-squares convex tensor regression to classification problems with a hinge loss and establish its asymptotic statistical properties. Based on a general convex decomposable penalty, the rate depends on both the intrinsic dimension and the Rademacher complexity of the class of linear functions of tensor predictors.
Original language | English |
---|---|
Article number | 9174817 |
Pages (from-to) | 3755-3760 |
Journal | IEEE Transactions on Neural Networks and Learning Systems |
Volume | 32 |
Issue number | 8 |
Online published | 24 Aug 2020 |
DOIs | |
Publication status | Published - Aug 2021 |
Research Keywords
- Empirical processes
- high-dimensional regression
- hinge loss
- intrinsic dimension
- Rademacher complexity
Fingerprint
Dive into the research topics of 'Learning Rate for Convex Support Tensor Machines'. Together they form a unique fingerprint.Projects
- 2 Finished
-
GRF: Low-rank tensor as a Dimension Reduction Tool in Complex Data Analysis
LIAN, H. (Principal Investigator / Project Coordinator)
1/01/20 → 28/11/24
Project: Research
-
GRF: Divide and Conquer in High-dimensional Statistical Models
LIAN, H. (Principal Investigator / Project Coordinator)
1/10/18 → 24/08/23
Project: Research