Singular Value Decomposition

تفكيك القيم المفردة

تحليل أي مصفوفة إلى جداء ثلاث مصفوفات: U (الاتجاهات اليسرى) وΣ (القيم المفردة) وV (الاتجاهات اليمنى)، مما يكشف بنيتها الرُّتبيّة.

Decomposing any matrix into a product of three matrices: U (left directions), Σ (singular values), and V (right directions), revealing its rank structure.

Also translated asSVD، التحليل إلى قيم مفردة

First appears in this corpus in: K-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation (2006)

Appears in these papers

K-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation2006 in the sky ✦
LoRA: Low-Rank Adaptation of Large Language Models2021 in the sky ✦
Matrix Factorization Techniques for Recommender Systems2009 in the sky ✦
Shampoo: Preconditioned Stochastic Tensor Optimization2018 in the sky ✦
Shampoo: Preconditioned Stochastic Tensor Optimization2018 in the sky ✦