New MPO-based framework enhances polynomial function approximation

Researchers have introduced a new framework called Multivariate Polynomial Optimization based on Matrix Product Operators (MPO)$^2$. This approach combines learned MPO feature embeddings with compact polynomial weight tensors to create efficient polynomial representations. The framework is feature order independent and can incorporate structured operators like projections and convolutions. In benchmarks for regression and classification, (MPO)$^2$ demonstrated improvements over existing tensor decomposition-based polynomial models, offering a flexible alternative for polynomial function approximation. AI

IMPACT Offers a more efficient and flexible approach to polynomial function approximation in machine learning tasks.

RANK_REASON The cluster contains an academic paper detailing a new method in machine learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv
• Canonical Polyadic Decompositions
• Hugging Face
• machine learning
• MPO$^2$
• signal processing
• Tensor-Train Decomposition

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →