Split Equality Coincidence Point Problems via Relaxed Cutter Pairs
We study the split equality coincidence point problem: given real Hilbert spaces \(H_1, H_2, H_3\), bounded linear operators \(A: H_1 \to H_3\), \(B: H_2 \to H_3\), \(T_2, S_2\), and (possibly nonlinear) operators \(T_1, S_1\), find \(x\) with \(T_1x = T_2x\) and \(y\) with \(S_1y = S_2y\) such that \(Ax = By\). The problem contains the split feasibility problem, the split equality problem and the split equality fixed point problem. We introduce \(\sigma\)-relaxed cutter pairs, a class which contains quasi-nonexpansive and demicontractive pairs, reduces to relaxed cutters when \(T_2 = I\), and does not require the nonlinear operator to factor through the linear one. For such pairs the coincidence set is the intersection of an explicit family of cutting half-spaces \(H_x\), computable from \(T_1, T_2\) and \(T_2^*\) alone, and the projection onto \(H_x\) is a norm-free cutter whose fixed point set is the coincidence set. Building on this, we propose a simultaneous method with self-adaptive step sizes, prove its weak convergence with an \(O(1/k)\) estimate for the residual, and, with a Halpern-type anchor, strong convergence to the projection of the anchor onto the solution set. Numerical experiments with singular \(T_2\) and non-factorizable pairs illustrate the results.
Authors
- Juan Martínez‐Moreno (ORCID: https://orcid.org/0000-0002-3340-2781)
Publication Details
- Journal
- Preprints.org
- Published
- 2026-11-13
- DOI
- https://doi.org/10.20944/preprints202611.1391.v1
- Primary Topic
- Optimization and Variational Analysis
- Type
- preprint