Pub Date : 2026-10-01Epub Date: 2026-02-23DOI: 10.1016/j.cam.2026.117457
Yanfang Yang, Lu Xiao
In this paper, we study higher-order Multiscale Finite Element Method (MsFEM) to solve linear elasticity equations with oscillating coefficients. Compared to the Finite Element Method (FEM), MsFEM can obtain the characteristics of fine scale in coarse mesh by using carefully designed multiscale basis functions. By using higher-order basis functions, better accuracy can be achieved. To further improve accuracy, several techniques are considered: oversampling methods and oscillatory boundary conditions (OBCs) are used to prevent the influence of boundary conditions on the construction of multiscale basis functions; the Petrov-Galerkin method, with the trial functions being multiscale basis functions and the test functions being polynomial functions. Numerical examples are presented to demonstrate the efficiency of the proposed methods.
{"title":"Higher-order multiscale finite element method for linear elasticity equations with oscillating coefficients","authors":"Yanfang Yang, Lu Xiao","doi":"10.1016/j.cam.2026.117457","DOIUrl":"10.1016/j.cam.2026.117457","url":null,"abstract":"<div><div>In this paper, we study higher-order Multiscale Finite Element Method (MsFEM) to solve linear elasticity equations with oscillating coefficients. Compared to the Finite Element Method (FEM), MsFEM can obtain the characteristics of fine scale in coarse mesh by using carefully designed multiscale basis functions. By using higher-order basis functions, better accuracy can be achieved. To further improve accuracy, several techniques are considered: oversampling methods and oscillatory boundary conditions (OBCs) are used to prevent the influence of boundary conditions on the construction of multiscale basis functions; the Petrov-Galerkin method, with the trial functions being multiscale basis functions and the test functions being polynomial functions. Numerical examples are presented to demonstrate the efficiency of the proposed methods.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117457"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386989","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-02-24DOI: 10.1016/j.cam.2026.117542
Hasanen A. Hammad , Tarek Aboelenen
This study investigates the demanding problem of proving the existence of solutions for a tripled system that couples quantum integral equations with quadratic integral equations. To tackle the intrinsic nonlinear and noncompact features of the model, we employ Petryshyn’s fixed-point theorem, a significant generalization of Darbo’s theorem formulated within the framework of measures of noncompactness. Based on this approach, we derive rigorous and verifiable existence conditions applicable to a broad class of quantum integral systems. The theoretical findings are supported by a comprehensive illustrative example that confirms the validity of the proposed criteria. In addition, we develop a constructive collocation method founded on barycentric interpolation and Jackson quadrature for q-integrals, and we verify the required assumptions within the same example. Numerical experiments are finally presented to confirm the practical applicability of the existence results and to demonstrate the accuracy, stability, and robustness of the proposed discretization scheme.
{"title":"Applying fixed point techniques for solving tripled system of quantum integral equations with numerical results","authors":"Hasanen A. Hammad , Tarek Aboelenen","doi":"10.1016/j.cam.2026.117542","DOIUrl":"10.1016/j.cam.2026.117542","url":null,"abstract":"<div><div>This study investigates the demanding problem of proving the existence of solutions for a tripled system that couples quantum integral equations with quadratic integral equations. To tackle the intrinsic nonlinear and noncompact features of the model, we employ Petryshyn’s fixed-point theorem, a significant generalization of Darbo’s theorem formulated within the framework of measures of noncompactness. Based on this approach, we derive rigorous and verifiable existence conditions applicable to a broad class of quantum integral systems. The theoretical findings are supported by a comprehensive illustrative example that confirms the validity of the proposed criteria. In addition, we develop a constructive collocation method founded on barycentric interpolation and Jackson quadrature for q-integrals, and we verify the required assumptions within the same example. Numerical experiments are finally presented to confirm the practical applicability of the existence results and to demonstrate the accuracy, stability, and robustness of the proposed discretization scheme.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117542"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386995","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-02-26DOI: 10.1016/j.cam.2026.117533
Hong Seo Ryoo
This paper illustrates with a small example that the algorithm proposed in [1] solves GLMP only locally, contrary to what is claimed in the paper that it is a global optimization algorithm.
{"title":"Comments on “An efficient algorithm for solving generalized linear multiplicative programming” by S. Liu and Y. Zhao","authors":"Hong Seo Ryoo","doi":"10.1016/j.cam.2026.117533","DOIUrl":"10.1016/j.cam.2026.117533","url":null,"abstract":"<div><div>This paper illustrates with a small example that the algorithm proposed in [1] solves GLMP only locally, contrary to what is claimed in the paper that it is a global optimization algorithm.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117533"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386901","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-02-26DOI: 10.1016/j.cam.2026.117481
Tao Han , Huancai Lu , D. Michael McFarland , Hanbo Jiang
This work presents a fusion of finite element analysis (FEA) and statistical energy analysis (SEA) for calculating the mid-frequency response in a stepped acoustic duct comprising two chambers, Chamber 1 (0.644 × 0.045 × 0.045 m3) coupled to Chamber 2 (0.17 × 0.02 × 0.02 m3), arranged coaxially and driven by a loudspeaker at the outer end of Chamber 1. The system was tested in a fully anechoic room over 178 to 11220 Hz. Based on modal-number classification, Chamber 1 acted as a mixed 1D/3D subsystem over 1415 to 8913 Hz, and so was modeled statistically, whereas Chamber 2 remained 1D and was treated deterministically. The subsystems were coupled by an interface with impedance dynamically condensed to a single-degree-of-freedom. Absorption coefficients measured in an impedance tube were used to derive internal loss factors. Source characterization combined a multi-microphone approach for frequencies below the first cutoff at 3811 Hz and a plane-wave approximation for higher frequencies. The SEA solution for Chamber 1 yielded interface pressure and volume velocity, while pre-computed transfer functions predicted pressures at probe positions in Chamber 2. Experiments showed maximum errors of 2.88 dB in Chamber 1 and 3.02 dB in Chamber 2. The proposed method required 1009 s of computation, and is thus about 20 × faster than full-system FEA, which required 20061 s.
{"title":"Fusion of FEA and SEA for mid-frequency acoustics in a variable-section duct: Impedance-based coupling and experimental validation","authors":"Tao Han , Huancai Lu , D. Michael McFarland , Hanbo Jiang","doi":"10.1016/j.cam.2026.117481","DOIUrl":"10.1016/j.cam.2026.117481","url":null,"abstract":"<div><div>This work presents a fusion of finite element analysis (FEA) and statistical energy analysis (SEA) for calculating the mid-frequency response in a stepped acoustic duct comprising two chambers, Chamber 1 (0.644 × 0.045 × 0.045 m<sup>3</sup>) coupled to Chamber 2 (0.17 × 0.02 × 0.02 m<sup>3</sup>), arranged coaxially and driven by a loudspeaker at the outer end of Chamber 1. The system was tested in a fully anechoic room over 178 to 11220 Hz. Based on modal-number classification, Chamber 1 acted as a mixed 1D/3D subsystem over 1415 to 8913 Hz, and so was modeled statistically, whereas Chamber 2 remained 1D and was treated deterministically. The subsystems were coupled by an interface with impedance dynamically condensed to a single-degree-of-freedom. Absorption coefficients measured in an impedance tube were used to derive internal loss factors. Source characterization combined a multi-microphone approach for frequencies below the first cutoff at 3811 Hz and a plane-wave approximation for higher frequencies. The SEA solution for Chamber 1 yielded interface pressure and volume velocity, while pre-computed transfer functions predicted pressures at probe positions in Chamber 2. Experiments showed maximum errors of 2.88 dB in Chamber 1 and 3.02 dB in Chamber 2. The proposed method required 1009 s of computation, and is thus about 20 × faster than full-system FEA, which required 20061 s.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117481"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386902","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-02-23DOI: 10.1016/j.cam.2026.117478
Jia-Zhi Feng, Cun-Qiang Miao
For the two-row block Kaczmarz method, we design a new deterministic row selection strategy. As a by-product, an explicit iterative expression for the two-row block Kaczmarz method is derived. We construct a fast deterministic two-row block Kaczmarz method and conduct its convergence analysis. Theoretical analysis and numerical experiments indicate that the proposed fast deterministic two-row block Kaczmarz method shows great superiority and robustness over some state-of-the art randomized block Kaczmarz methods.
{"title":"On fast deterministic two-row block Kaczmarz method for solving consistent linear systems","authors":"Jia-Zhi Feng, Cun-Qiang Miao","doi":"10.1016/j.cam.2026.117478","DOIUrl":"10.1016/j.cam.2026.117478","url":null,"abstract":"<div><div>For the two-row block Kaczmarz method, we design a new deterministic row selection strategy. As a by-product, an explicit iterative expression for the two-row block Kaczmarz method is derived. We construct a fast deterministic two-row block Kaczmarz method and conduct its convergence analysis. Theoretical analysis and numerical experiments indicate that the proposed fast deterministic two-row block Kaczmarz method shows great superiority and robustness over some state-of-the art randomized block Kaczmarz methods.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117478"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147387135","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-02-22DOI: 10.1016/j.cam.2026.117488
Bin Tang, Jinkui Liu, Ting Liu, Shuo Liang, Zhitong Yang
In this paper, we propose a multi-step derivative-free projection method for solving large-scale nonlinear equations with convex constraints. The first direction is constructed as a modified steepest descent direction, while the second direction is derived by equating the spectral gradient direction with the classical conjugate gradient direction, incorporating a convex combination of the HS and DY parameters. These dual-direction mechanisms synergistically enhance the efficiency of the search process. Notably, the global convergence of the proposed method does not rely on the Lipschitz continuity assumption. Furthermore, both asymptotic and non-asymptotic convergence rates are established in terms of iteration complexity. Numerical experiments validate its efficacy in addressing large-scale nonlinear equations and sparse signal recovery problems.
{"title":"A multi-step derivative-free projection method for nonlinear equations with application to sparse signal recovery","authors":"Bin Tang, Jinkui Liu, Ting Liu, Shuo Liang, Zhitong Yang","doi":"10.1016/j.cam.2026.117488","DOIUrl":"10.1016/j.cam.2026.117488","url":null,"abstract":"<div><div>In this paper, we propose a multi-step derivative-free projection method for solving large-scale nonlinear equations with convex constraints. The first direction is constructed as a modified steepest descent direction, while the second direction is derived by equating the spectral gradient direction with the classical conjugate gradient direction, incorporating a convex combination of the HS and DY parameters. These dual-direction mechanisms synergistically enhance the efficiency of the search process. Notably, the global convergence of the proposed method does not rely on the Lipschitz continuity assumption. Furthermore, both asymptotic and non-asymptotic convergence rates are established in terms of iteration complexity. Numerical experiments validate its efficacy in addressing large-scale nonlinear equations and sparse signal recovery problems.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117488"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147387137","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-02-20DOI: 10.1016/j.cam.2026.117450
Xiongxiong Du, Xiaoling Han
In this study, we explore the spatiotemporal dynamics of a discrete time-space predator-prey system with prey refuge and cross-diffusion. Through stability and bifurcation analyses, the conditions for the formation of Turing patterns are derived, and three nonlinear mechanisms for pattern formation are found, namely, pure Turing instability, flip-Turing instability and Neimark-Sacker-Turing instability. Numerical simulations have unveiled the rich dynamics within discrete predator-prey model. In spatially homogeneous conditions, it exhibits stable homogeneous steady states, homogeneous periodic, quasi-periodic and chaotic oscillatory states. In spatially inhomogeneous conditions, various prey patterns are described, including spots, stripes, labyrinths, spirals, phobic patterns and many intermediate patterns. These richer nonlinear dynamical characteristics contribute to a deeper understanding of the complex pattern formation in spatially diffusion discrete predator-prey systems.
{"title":"Analysis of spatiotemporal complexity in a discrete time-space predator-prey system with prey refuge and cross-diffusion","authors":"Xiongxiong Du, Xiaoling Han","doi":"10.1016/j.cam.2026.117450","DOIUrl":"10.1016/j.cam.2026.117450","url":null,"abstract":"<div><div>In this study, we explore the spatiotemporal dynamics of a discrete time-space predator-prey system with prey refuge and cross-diffusion. Through stability and bifurcation analyses, the conditions for the formation of Turing patterns are derived, and three nonlinear mechanisms for pattern formation are found, namely, pure Turing instability, flip-Turing instability and Neimark-Sacker-Turing instability. Numerical simulations have unveiled the rich dynamics within discrete predator-prey model. In spatially homogeneous conditions, it exhibits stable homogeneous steady states, homogeneous periodic, quasi-periodic and chaotic oscillatory states. In spatially inhomogeneous conditions, various prey patterns are described, including spots, stripes, labyrinths, spirals, phobic patterns and many intermediate patterns. These richer nonlinear dynamical characteristics contribute to a deeper understanding of the complex pattern formation in spatially diffusion discrete predator-prey systems.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117450"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386624","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-02-21DOI: 10.1016/j.cam.2026.117479
Ke-Yu Zhu , Gui-Xian Tian , Shu-Yu Cui
Given a graph G with vertex set and a graph H of order n2, the vertex complemented corona, denoted by , is the graph produced by copying H n1 times, with the ith copy of H corresponding to the vertex vi, and then adding edges between any vertex in V(G)∖{vi} and any vertex of the ith copy of H. The present article deals with quantum state transfer of vertex complemented coronas concerning signless Laplacian matrix. Our research investigates conditions in which signless Laplacian perfect state transfer exists or not on vertex complemented coronas. Additionally, we also provide some mild conditions for the class of graphs under consideration that allow signless Laplacian pretty good state transfer.
{"title":"Signless Laplacian state transfer on vertex complemented coronas","authors":"Ke-Yu Zhu , Gui-Xian Tian , Shu-Yu Cui","doi":"10.1016/j.cam.2026.117479","DOIUrl":"10.1016/j.cam.2026.117479","url":null,"abstract":"<div><div>Given a graph <em>G</em> with vertex set <span><math><mrow><mi>V</mi><mrow><mo>(</mo><mi>G</mi><mo>)</mo></mrow><mo>=</mo><mrow><mo>{</mo><msub><mi>v</mi><mn>1</mn></msub><mo>,</mo><msub><mi>v</mi><mn>2</mn></msub><mo>,</mo><mo>…</mo><mo>,</mo><msub><mi>v</mi><msub><mi>n</mi><mn>1</mn></msub></msub><mo>}</mo></mrow></mrow></math></span> and a graph <em>H</em> of order <em>n</em><sub>2</sub>, the vertex complemented corona, denoted by <span><math><mrow><mi>G</mi><mover><mo>∘</mo><mo>˜</mo></mover><mi>H</mi></mrow></math></span>, is the graph produced by copying <em>H n</em><sub>1</sub> times, with the <em>i</em>th copy of <em>H</em> corresponding to the vertex <em>v<sub>i</sub></em>, and then adding edges between any vertex in <em>V</em>(<em>G</em>)∖{<em>v<sub>i</sub></em>} and any vertex of the <em>i</em>th copy of <em>H</em>. The present article deals with quantum state transfer of vertex complemented coronas concerning signless Laplacian matrix. Our research investigates conditions in which signless Laplacian perfect state transfer exists or not on vertex complemented coronas. Additionally, we also provide some mild conditions for the class of graphs under consideration that allow signless Laplacian pretty good state transfer.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117479"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386625","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-02-22DOI: 10.1016/j.cam.2026.117477
Omar El Ogri , Jaouad EL-Mekkaoui , Mohamed Benslimane , Amal Hjouji
Image analysis is a classic and commonplace task in the field of computer vision, widely applied over the past decade. Many existing methods in the literature are designed for signal and image analysis using the moments method show that most Dual Hahn moments applications are based on orthogonal polynomials of low order (n ≤ 128). However, the computation of high-order Dual Hahn polynomials remains highly constrained. Consequently, the primary objective of this study is to introduce two novel, stable, and efficient algorithms specifically designed for the computation of high-order Dual Hahn moments. The two algorithms rely on recently developed recurrence relations and the Gram-Schmidt Process (GSP), which take into account both the variable and order , removing the terms responsible for numerical fluctuations and excessive computation time, especially at high orders. The GSP is then commonly used to correct numerical instability during the calculation of high-order Discrete Orthogonal Dual Hahn Polynomials (DODHPs). These algorithms accelerate the implementation of DODHPs and ensure the numerical stability of orthogonal moments up to the final order through an analysis of the coefficient distribution within the polynomial matrix. An efficient method has also been developed to expedite the reconstruction time of large size 1D signals. To evaluate the proposed algorithms, we present several experimental tests on sets of signals and images. In this context, we evaluate our algorithms for compression and reconstruction of large 1D and 2D signals. Then, in recognition, we used our descriptor vector based on the proposed algorithms for image feature extraction, as well as the deep learning method DNN for image classification and prediction. These results demonstrate that the proposed algorithms for the speed and stability of large-size signals and 2D images outperform conventional methods and other types of existing moments.
{"title":"An efficient computational high-order Dual Hahn polynomials approach for reconstruction, compression, and recognition of large-size signals using machine learning","authors":"Omar El Ogri , Jaouad EL-Mekkaoui , Mohamed Benslimane , Amal Hjouji","doi":"10.1016/j.cam.2026.117477","DOIUrl":"10.1016/j.cam.2026.117477","url":null,"abstract":"<div><div>Image analysis is a classic and commonplace task in the field of computer vision, widely applied over the past decade. Many existing methods in the literature are designed for signal and image analysis using the moments method show that most Dual Hahn moments applications are based on orthogonal polynomials of low order (<em>n</em> ≤ 128). However, the computation of high-order Dual Hahn polynomials remains highly constrained. Consequently, the primary objective of this study is to introduce two novel, stable, and efficient algorithms specifically designed for the computation of high-order Dual Hahn moments. The two algorithms rely on recently developed recurrence relations and the Gram-Schmidt Process (GSP), which take into account both the variable <span><math><mi>s</mi></math></span> and order <span><math><mi>n</mi></math></span>, removing the terms responsible for numerical fluctuations and excessive computation time, especially at high orders. The GSP is then commonly used to correct numerical instability during the calculation of high-order Discrete Orthogonal Dual Hahn Polynomials (DODHPs). These algorithms accelerate the implementation of DODHPs and ensure the numerical stability of orthogonal moments up to the final order through an analysis of the coefficient distribution within the polynomial matrix. An efficient method has also been developed to expedite the reconstruction time of large size 1D signals. To evaluate the proposed algorithms, we present several experimental tests on sets of signals and images. In this context, we evaluate our algorithms for compression and reconstruction of large 1D and 2D signals. Then, in recognition, we used our descriptor vector based on the proposed algorithms for image feature extraction, as well as the deep learning method DNN for image classification and prediction. These results demonstrate that the proposed algorithms for the speed and stability of large-size signals and 2D images outperform conventional methods and other types of existing moments.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117477"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386628","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-02-24DOI: 10.1016/j.cam.2026.117469
Xiangli Li , Zhiling Wang , Binglan Li
In this paper, based on the excellent properties of Newton method, and the motivation of sufficient descent condition, we propose a new spectral hybrid conjugate gradient method. By using the secant line condition, the appropriate combination weight parameter is calculated. The spectral parameter is obtained under the rule of sufficient descent for search direction without any line search. Using the Wolfe line search, we prove the global convergence of the proposed method. Finally, numerical results show that the proposed method is effective.
{"title":"A improved spectral hybrid conjugate gradient method for unconstrained optimization","authors":"Xiangli Li , Zhiling Wang , Binglan Li","doi":"10.1016/j.cam.2026.117469","DOIUrl":"10.1016/j.cam.2026.117469","url":null,"abstract":"<div><div>In this paper, based on the excellent properties of Newton method, and the motivation of sufficient descent condition, we propose a new spectral hybrid conjugate gradient method. By using the secant line condition, the appropriate combination weight parameter is calculated. The spectral parameter is obtained under the rule of sufficient descent for search direction without any line search. Using the Wolfe line search, we prove the global convergence of the proposed method. Finally, numerical results show that the proposed method is effective.</div></div>","PeriodicalId":50226,"journal":{"name":"Journal of Computational and Applied Mathematics","volume":"484 ","pages":"Article 117469"},"PeriodicalIF":2.6,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号