{"id":"arxiv/2608.01273","url":"https://arxiv.org/abs/2608.01273","tags":["math.NA"],"score":0,"title":"On a Class of Block-Regularized Preconditioned Iterative Methods for Indefinite Least-Squares Problems","domain":"math.NA","source":"arxiv","authors":["Pinki Khatun"],"reviews":0,"subject":"math.NA","summary":"This paper proposes a novel block-regularized splitting (BRS) framework for the efficient solution of indefinite least-squares (ILS) problems. Based on a block-wise regularization strategy incorporated into a matrix splitting scheme, we develop a BRS iterative method together with an effective BRS preconditioner. The convergence of the proposed iterative method is rigorously analyzed, and spectral bounds for the BRS-preconditioned matrix are established. To further enhance computational performance, we introduce a relaxed BRS (RBRS) preconditioner, which provides improved spectral properties…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-08-02","reactions":{}}