arXiv cs.AI· Albert Saiapin, Kim Batselier·· 12 小时前AI 评分20
GB-CPKM：用梯度提升实现张量网络核机的高效秩发现
Gradient Boosting for Efficient Rank Discovery in Tensor Network Kernel Machines
AI 导读
研究者提出 GB-CPKM，一种基于 CP 张量网络参数化的梯度提升框架，通过顺序添加固定秩学习器逐步提升模型容量，以验证集表现选择集成规模，替代显式秩搜索。在真实回归基准上，GB-CPKM 训练速度比基于交叉验证的传统秩选择方法快至多 6 倍，峰值内存少至多 2 倍，预测性能相当。
正文 · 原文
Abstract:Tensor network kernel machines (TNKMs) combine nonlinear features with low-rank tensor representations to model complex relationships efficiently. However, selecting an appropriate tensor rank remains challenging, as conventional rank-search methods require training multiple candidate models, increasing computational costs. In this work, we focus on a canonical polyadic (CP) tensor network parameterization and propose GB-CPKM, a gradient boosting framework that builds a higher-capacity predictor by sequentially adding fixed-rank learners. This approach provides a new perspective on rank selection by incrementally increasing model capacity, with ensemble size selected based on validation performance rather than through explicit rank search. Experiments on real-world regression benchmarks show that GB-CPKM trains up to 6x faster and uses up to 2x less peak memory than a conventional rank-selection method based on cross-validation, while achieving comparable predictive performance. These results highlight the potential of gradient boosting to improve the computational efficiency and scalability of tensor-network learning.
| Comments: | 13 pages, 3 figures, 1 table. Code available at: this https URL |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.08093 [cs.LG] |
| (or arXiv:2611.08093v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2611.08093 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Albert Saiapin [view email]
[v1]
Thu, 5 Nov 2026 14:36:46 UTC (5,754 KB)
来源：arXiv cs.AI · arxiv.org