{"id":"arxiv/2606.31003","url":"https://arxiv.org/abs/2606.31003","tags":["math.NA","cs.MS"],"score":0,"title":"Ricci-Notation Tensor Framework for Numerical Algebraic Geometry via Any-Degree Unitary-Triangular Factorization","domain":"math.NA","source":"arxiv","authors":["Dileepan Joseph"],"reviews":0,"subject":"math.NA","summary":"The unitary-triangular (QR) factorization of linear algebra may be used to robustly and efficiently solve a linear system. Toward a comparable numerical method to solve a polynomial system of higher degree, this paper proposes an any-degree unitary-triangular (Qr) factorization, which for a degree-one system reduces to the QR factorization. The work develops a tensor framework, i.e., codesigned algebra and software, where polynomial system coefficients are represented by a vector-shaped sparse tensor, a multidimensional array whose number of Ricci-notation indices, called the tensor degree,…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-06-30","reactions":{}}