What the study found
A new hybrid conjugate gradient method was proposed for unconstrained optimization and was reported to be globally convergent under exact line search.
Why the authors say this matters
The authors state that conjugate gradient methods are important for unconstrained optimization, especially for large-scale problems, and the findings indicate the new hybrid method may be more efficient than previous methods.
What the researchers tested
The researchers proposed a hybrid conjugate gradient method that combines an earlier version of Polak-Ribiere and Polyak (PRP) with a recent modification of the Mouiyad Bani Yousef (MMR) method. They tested two combinations of the new hybrid method and the PRP conjugate gradient method using numerical experiments.
What worked and what didn't
The proposed method was proved globally convergent under exact line search, and the numerical tests supported this result. The numerical performance of the new hybrid method was reported to be more efficient compared with previous conjugate gradient methods, and one hybrid version was especially effective for the test problems.
What to keep in mind
The abstract does not describe detailed limitations, and the results are presented from numerical tests on the reported test problems.
Key points
- The paper proposes a new hybrid conjugate gradient method for unconstrained optimization.
- The hybrid method combines an earlier PRP method with a modified MMR method.
- The method was proved globally convergent under exact line search.
- Numerical tests supported the convergence result.
- The new hybrid method was reported to be more efficient than previous conjugate gradient methods.
Disclosure
- Research title:
- Hybrid PRP-MMR method is globally convergent under exact line search
- Authors:
- Mouiyad Bani Yousef, Mustafa Mamat, Mohd Rivaie
- Publication date:
- 2026-04-20
- OpenAlex record:
- View
- Image credit:
- Photo by Anrita1705 on Pixabay · Pixabay License
Get the weekly research newsletter
Stay current with peer-reviewed research without reading academic papers — one filtered digest, every Friday.