Loading [MathJax]/jax/output/SVG/jax.js
Research article

Time-adaptive Lagrangian variational integrators for accelerated optimization

  • A variational framework for accelerated optimization was recently introduced on normed vector spaces and Riemannian manifolds in [1] and [2]. It was observed that a careful combination of time-adaptivity and symplecticity in the numerical integration can result in a significant gain in computational efficiency. It is however well known that symplectic integrators lose their near-energy preservation properties when variable time-steps are used. The most common approach to circumvent this problem involves the Poincaré transformation on the Hamiltonian side, and was used in [3] to construct efficient explicit algorithms for symplectic accelerated optimization. However, the current formulations of Hamiltonian variational integrators do not make intrinsic sense on more general spaces such as Riemannian manifolds and Lie groups. In contrast, Lagrangian variational integrators are well-defined on manifolds, so we develop here a framework for time-adaptivity in Lagrangian variational integrators and use the resulting geometric integrators to solve optimization problems on vector spaces and Lie groups.

    Citation: Valentin Duruisseaux, Melvin Leok. Time-adaptive Lagrangian variational integrators for accelerated optimization[J]. Journal of Geometric Mechanics, 2023, 15(1): 224-255. doi: 10.3934/jgm.2023010

    Related Papers:

    [1] Xiaoli Zhang, Ziying Yu . Pathological analysis of hesperetin-derived small cell lung cancer by artificial intelligence technology under fiberoptic bronchoscopy. Mathematical Biosciences and Engineering, 2021, 18(6): 8538-8558. doi: 10.3934/mbe.2021423
    [2] Ming Zhang, Yingying Zhou, Yanli Zhang . High Expression of TLR2 in the serum of patients with tuberculosis and lung cancer, and can promote the progression of lung cancer. Mathematical Biosciences and Engineering, 2020, 17(3): 1959-1972. doi: 10.3934/mbe.2020104
    [3] Bijiong Wang, Yaodong Tang, Biyun Yu, Di Gui, Hui Xu . Expression of autophagy-related factor p62 for lung cancer diagnosis and prognosis: A systematic review and meta-analysis. Mathematical Biosciences and Engineering, 2019, 16(6): 6805-6821. doi: 10.3934/mbe.2019340
    [4] Gaozhong Sun, Tongwei Zhao . Lung adenocarcinoma pathology stages related gene identification. Mathematical Biosciences and Engineering, 2020, 17(1): 737-746. doi: 10.3934/mbe.2020038
    [5] Kunpeng Li, Zepeng Wang, Yu Zhou, Sihai Li . Lung adenocarcinoma identification based on hybrid feature selections and attentional convolutional neural networks. Mathematical Biosciences and Engineering, 2024, 21(2): 2991-3015. doi: 10.3934/mbe.2024133
    [6] Guohai Wang, Zhiyuan Hu . Icotinib inhibits proliferation and epithelial-mesenchymal transition of non-small cell lung cancer A549 cells. Mathematical Biosciences and Engineering, 2019, 16(6): 7707-7718. doi: 10.3934/mbe.2019386
    [7] Andreas Wagner, Pirmin Schlicke, Marvin Fritz, Christina Kuttler, J. Tinsley Oden, Christian Schumann, Barbara Wohlmuth . A phase-field model for non-small cell lung cancer under the effects of immunotherapy. Mathematical Biosciences and Engineering, 2023, 20(10): 18670-18694. doi: 10.3934/mbe.2023828
    [8] Yong Ding, Jian-Hong Liu . The signature lncRNAs associated with the lung adenocarcinoma patients prognosis. Mathematical Biosciences and Engineering, 2020, 17(2): 1593-1603. doi: 10.3934/mbe.2020083
    [9] Yongxing Li, Jianye Yang, Yufen Xu, Ming Zhang, Xiaoping Zhang, Wenyu Chen, Xiaodong Lv . A meta-analysis of the comparing of the first-generation and next-generation TKIs in the treatment of NSCLC. Mathematical Biosciences and Engineering, 2019, 16(5): 5687-5696. doi: 10.3934/mbe.2019283
    [10] Lingling Li, Mengyao Shao, Xingshi He, Shanjing Ren, Tianhai Tian . Risk of lung cancer due to external environmental factor and epidemiological data analysis. Mathematical Biosciences and Engineering, 2021, 18(5): 6079-6094. doi: 10.3934/mbe.2021304
  • A variational framework for accelerated optimization was recently introduced on normed vector spaces and Riemannian manifolds in [1] and [2]. It was observed that a careful combination of time-adaptivity and symplecticity in the numerical integration can result in a significant gain in computational efficiency. It is however well known that symplectic integrators lose their near-energy preservation properties when variable time-steps are used. The most common approach to circumvent this problem involves the Poincaré transformation on the Hamiltonian side, and was used in [3] to construct efficient explicit algorithms for symplectic accelerated optimization. However, the current formulations of Hamiltonian variational integrators do not make intrinsic sense on more general spaces such as Riemannian manifolds and Lie groups. In contrast, Lagrangian variational integrators are well-defined on manifolds, so we develop here a framework for time-adaptivity in Lagrangian variational integrators and use the resulting geometric integrators to solve optimization problems on vector spaces and Lie groups.



    Lung cancer is the leading cause of cancer-related death in men and the second leading cause (after breast cancer) in women worldwide; more than 1 million deaths are attributed to this disease annually [1,2,3]. Non-small-cell-lung cancer (NSCLC) accounts for approximately 80–85% of all lung malignancies, and presents as locally advanced NSCLC in approximately 25–30% of cases and as metastatic disease in approximately 40–50% of cases [4]. The preferred treatment for lung cancer is surgical resection; however, many patients with advanced NSCLC still experience relapse despite complete excision; the 5-year survival for patients who only undergo surgical treatment is less than 25% [5].

    Platinum-based chemotherapy is the mainstay of treatment for patients who are ineligible for surgical treatment. Although patients may benefit from chemotherapy in the short term, their response rates are ultimately disappointing as most eventually experience recurrence. The screening, diagnosis, and treatment of NSCLC have made considerable progress over the past decade. Of note is the discovery of targetable molecular markers such as epidermal growth factor receptor (EGFR). EGFR tyrosine kinase inhibitors are the first-line treatments for patients with advanced NSCLC whose tumors are positive for EGFR mutations; these inhibitors increase treatment efficacy, decrease toxic effects, and improve outcomes [6,7,8]. However, EGFR mutations are found only in 10–30% of patients with NSCLC [9], and the initial treatment for patients with advanced NSCLC without genetic mutations remains platinum-based doublet chemotherapy [10]. Patients in whom second or later-line treatments fail have no effective treatment options available, although many are administered chemotherapy in practice [11].

    Angiogenesis is essential for tumor growth and metastasis [12]. Vascular endothelial growth factor (VEGF), which activates the VEGF receptor (VEGFR), has long been known to have a critical role in promoting tumor angiogenesis [13]. Hence, VEGF and VEGFR are important targets in cancer therapy, and VEGF inhibition has been proven to be effective against many solid tumors [14,15,16]. The anti-VEGF antibody bevacizumab has been shown to significantly improve PFS and overall survival (OS) in patients with advanced NSCLC in several studies [17,18].

    Apatinib, a novel orally bioavailable small molecule inhibitor of VEGFR-2, could inhibit VEGF-stimulated endothelial cell migration and proliferation and decrease tumor microvascular density, thus inhibiting the growth of tumors. This might lead to prolonging the survival time of patients. It has been shown to be efficient and safe for the treatment of multiple tumor types, such as breast and gastric cancers [14,16,19]. Apatinib also be used in patients with advanced NSCLC for second or later-line treatments [20]. When chemotherapy and/or targeted therapy fail, it might become a new option. However, the efficacy and safety of Apatinib in this case was not clear. The meta-analysis described herein was performed to assess the efficacy and toxicity of Apatinib in patients with advanced NSCLC.

    In November 2017, we searched the PubMed, EMBASE, Cochrane Library, China National Knowledge Internet (CNKI), and Wanfang database for randomized controlled trials using the keywords ("NSCLC" OR "Carcinoma, Non-Small Cell Lung") AND ("Apatinib" OR "YN968D1") without language or year restrictions. The reference lists of recovered articles were also searched for related original and review articles to screen for additional trials.

    Two investigators independently assessed candidate articles for inclusion. Any disagreements between the 2 assessments were resolved by discussion with a third investigator. The relevant studies were manually and carefully selected based on the following criteria: (1) randomized controlled trial, (2) patients had pathologically confirmed NSCLC, (3) studies compared the efficacy and toxicity of apatinib with a placebo or chemotherapy, we defined the treatment group without apatinib as the control group, and (4) studies contained sufficient data for extraction. The Cochrane collaboration's tool was used to evaluate the risk of bias in the studies [21].

    The same 2 investigators independently extracted the necessary information from each of the included articles, then cross-checked the information. Discrepancies were settled by discussion with a third investigator.

    The following information were collected from each study unless otherwise not available: First author, year of publication, number of patients, patient characteristics (such as sex), and outcomes (including median PFS, ORR, and DCR values; HRs for PFS and their 95% CIs; and adverse events such as hand-foot syndrome and hypertension).

    RevMan 5.3 software (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) was used to perform statistical analysis. The Engauge Digitizer 4.1 software was used to extract survival data from the Kaplan-Meier curves [22,23]. The pooled HRs for PFS, RRs for ORRs and DCRs, and ORs for different adverse events were calculated. The statistical heterogeneity between trials was evaluated by the Q-statistic [24]. A Q Statistic P-value < 0.1 or an I2 > 50% indicated significant statistical heterogeneity between studies, in which case a random-effects model was used to analyze the data. Otherwise, a fixed-effects model was used. Subetaoup analysis was used to reduce heterogeneity. A statistical test with a P-value < 0.05 was considered significant. An HR > 1 indicated a greater death or progression rate upon treatment with apatinib, an RR > 1 indicated a greater overall response, and an OR > 1 indicated greater toxicities due to therapy. All P-values were 2-sided, and all CIs had 2-sided probability coverages of 95%.

    Fifty-eight candidate articles were identified by systematically searching the relevant databases using our search strategy, of which 5 qualified articles describing studies that compared the efficacy and toxicity of apatinib to controls in patients with advanced NSCLC were included based on our criteria (Figure 1) [20,25,26,27,28]. A study was excluded because the data was incomplete and ambiguous [29]. All the articles were written in Chinese, and all studies were conducted in China. The characteristics of these studies are shown in Table 1. The 5 clinical studies comprised 413 patients; the number of patients per study ranged from 39 to 135. Results of the analyses of risk of bias in the studies are shown in Figure 2.

    Figure 1.  Diagram of the literature search and trial selection process in the meta-analysis.
    Table 1.  The characeristics of included studies.
    Study Year Comparison arms Participants Median PFS (mo) ORR (%) DCR (%)
    Li Zhang 2012 Apatinib 90 (44) 4.7 12.2 68.9
    Placebo 45 (23) 1.9 0 24.4
    Jinling Bi 2017 Apatinib 35 (-) - 17.1 65.7
    Docetaxel 30 (-) - 10.0 46.7
    Erhong Chen 2017 Apatinib 42 (26) - 50.0 95.2
    Paclitaxel 42 (25) - 33.3 81.0
    Yunjie Guo 2017 Apatinib plus docetaxel 19 (10) 5.6 31.6 57.9
    Docetaxel 20 (10) 3.0 10.0 30.0
    Yongjing Liu 2017 Apatinib plus paclitaxel plus cisplatin 45 (-) - 35.6 55.6
    Paclitaxel plus cisplatin 45 (-) - 17.8 33.3

     | Show Table
    DownLoad: CSV
    Figure 2.  Diagram of risk of bias in the studies.

    Only 2 of the studies investigated PFS rates, with no significant heterogeneity between them (P = 0.34, I2 = 0%). We used the fixed-effects model to estimate the pooled hazard ratio (HR) for PFS, which showed that therapy with apatinib significantly improved PFS in patients with advanced NSCLC (HR: 0.32, 95% confidence interval (CI): 0.21–0.48, P < 0.00001, Figure 3). Apatinib was used as a third- or later-line treatment in Li Zhang's study [20], so placebo was used as a control group. While in the study of Yunjie Guo [27], apatinib was used as a second-line treatment, it was not appropriate to use a placebo as the control group. However, no matter what treatment was used as the control group, apatinib benefited patients from PFS

    Figure 3.  Forest plot of comparison for PFS between apatinib and control group in advanced NSCLC.

    All 5 studies investigated the ORR; there was no significant difference between them (P = 0.53, I2 = 0%). The total pooled RR for ORR, which was calculated using the fixed-effects model, was 2.03 (95% CI: 1.36–3.01, P = 0.0005), and was significantly higher in the apatinib group than the control group (Figure 4), indicating that apatinib improved the ORR. Subetaoup analysis revealed that the ORRs of the apatinib and chemotherapy groups were not significantly different (P = 0.09). However, the ORRs were significantly higher in the apatinib plus chemotherapy group than in the chemotherapy-only group (P = 0.02). When using placebo as the control group, ORR was lower than other studies using chemotherapy as the control group. The treatment group containing apatinib effectively improved ORR.

    Figure 4.  Forest plot of comparison for ORR between apatinib and control group in advanced NSCLC.

    All 5 studies investigated the DCR; the differences between the studies were significant (P = 0.0003, I2 = 81%).The total pooled RR for the DCR (calculated using the random-effects model) was 1.66 (95% CI: 1.07–2.57, P = 0.02), and was significantly higher in apatinib-treated patients than in the control group (Figure 5). Subetaoup analysis was used to reduce heterogeneity. The differences between the studies' apatinib vs. chemotherapy-alone and apatinib plus chemotherapy vs. chemotherapy-alone groups were not significant (P = 0.38, I2 = 0%; and P = 0.75, I2 = 0%; respectively). The corresponding pooled RRs for DCR were 1.20 (95% CI: 1.03–1.40, P = 0.02) and 1.74 (95% CI: 1.15–2.63, P = 0.009), respectively, demonstrating that apatinib therapy significantly improved the DCR. DCR was lower in the placebo control group than that in chemotherapy control group. The treatment group with apatinib effectively improved DCR.

    Figure 5.  Forest plot of comparison for DCR between apatinib and control group in advanced NSCLC.

    Gastrointestinal reaction, hand-foot syndrome, hypertension, thrombocytopenia, and hypoleucocytosis were the most common adverse events. The pooled OR was 1.34 (95% CI: 0.57–3.17) (Figure 6), indicating less toxicity in the control group than in the apatinib group; however, the difference was not statistically significant (P = 0.5). There was high heterogeneity (I2 = 81%) between the apatinib and control groups; therefore, a random-effects model was employed. For gastrointestinal reaction and hypoleucocytosis, the pooled OR were 0.66 (95% CI: 0.13–3.41) and 0.41 (95% CI: 0.04–3.80), respectively, indicating more gastrointestinal reaction and hypoleucocytosis in the control group than in the apatinib group; however, the difference was not statistically significant (P = 0.62 and P = 0.43). The heterogeneity (I2 = 84% and I2 = 89%) between the apatinib and control groups was significant. For hypertension, hand-foot syndrome, thrombocytopenia, the pooled OR were 11.83 (95% CI: 2.60–53.84), 3.77 (95% CI: 0.23–60.72) and 1.60 (95% CI: 0.39–6.56), respectively, indicating less hypertension, hand-foot syndrome, thrombocytopenia in the control group than in the apatinib group; The difference was not statistically significant (P = 0.51 and P = 0.35) for hand-foot syndrome and thrombocytopenia and the difference was statistically significant (P = 0.001) for hypertension. There was high heterogeneity (I2 = 83% and I2 = 67%) between the apatinib and control groups for hand-foot syndrome and thrombocytopenia. The heterogeneity (I2 = 18%) between the apatinib and control groups for hypertension was not significant. Subetaoup analysis showed that the differences between the studies in each category were statistically significant except for hypertension (I2 = 18%). The incidences of hypertension, hand-foot syndrome, and thrombocytopenia in the apatinib group were higher than those in the control group, although only in the hypertension group was the difference significant (OR: 11.83, 95% CI: 2.60–53.84, P = 0.001). The incidences of gastrointestinal reactions and hypoleucocytosis in the apatinib group were fewer than in the control group, although the differences were not statistically significant.

    Figure 6.  Forest plot of comparison for toxicity between apatinib and control group in advanced NSCLC.

    To our knowledge, ours is the first meta-analysis to assess the clinical benefit of apatinib as a post-first-line treatment in patients with advanced NSCLC. Our results demonstrated that apatinib therapy for advanced NSCLC as a second-or-subsequent-line treatment provides a substantial clinical benefit. Apatinib may therefore be a novel choice for patients with advanced NSCLC who fail second- or higher-line treatments, given that no standard treatment regimens are yet recommended for such patients.

    Apatinib is a first-generation oral small molecule VEGFR inhibitor that has been independently researched and developed in China; its main target is VEGFR-2 [30]. Several studies have confirmed that blocking VEGFR-2 effectively inhibits angiogenesis [31,32], and apatinib has been approved by the China Food and Drug Administration for treating advanced or metastatic chemorefractory gastric cancer [14]. Several studies had been reported for the application of apatinib in patients with advanced NSCLC [33,34,35]. To date, the vast majority of studies involving apatinib for advanced NSCLC were conducted in China; therefore, all the papers evaluated in this meta-analysis were authored in China. Because we systematically assessed the efficacy and adverse reactions of apatinib in patients with advanced NSCLC in related randomized trials, the results of our study may be applicable to patients in other countries.

    This meta-analysis covered all the currently published randomized studies in which apatinib was used as a post-first-line treatment for advanced NSCLC. The results confirmed that treatment with apatinib provided substantial clinical benefits for patients in term of PFS, ORR, and DCR. The ORR of the apatinib group was higher than that of the chemotherapy group, although the difference was not statistically significant. However, the difference in the ORRs of the apatinib plus chemotherapy vs. chemotherapy-only groups was significant. In terms of DCR, the differences between the studies in each subetaoup were not statistically significant (I2 = 0%), and all showed that apatinib provides a substantial benefit to patients with advanced NSCLC.

    Overall survival is usually the endpoint for evaluating clinical benefit, whereas PFS is considered a surrogate endpoint. However, not all investigations included in our meta-analysis reported overall survival rates, assessment of which was therefore absent. The significant improvements in PFS, ORR, and DCR ought to be meaningful to the patients; therefore, a fair recommendation that apatinib be used to treat patients with advanced NSCLC is reasonable.

    In general, the adverse events were similar among patients in the apatinib and control groups (P = 0.50), although there was high heterogeneity between the studies. The differences between the hypertension subetaoups in the studies were not statistically significant, even though the incidence of hypertension in the apatinib group was significantly higher than that in the control group. Hence, the results of other subetaoup analyses should be considered with caution.

    There were several limitations in this study. First, the different methods of treatment contributed to increased clinical heterogeneity. Second, our meta-analysis was based on data extracted from published literature, not on individual patient data. Third, the quantity and quality of research included in this meta-analysis were limited; more high-quality randomized controlled trials are warranted to investigate the efficacy of apatinib against advanced NSCLC.

    Apatinib is an effective choice for patients with advanced NSCLC, and produces a clinically meaningful and statistically significant improvement in PFS, ORR, and DCR. Adverse reactions associated with apatinib were acceptable. Therapy with apatinib-based regimens is suggested for second- or later-line treatment of advanced NSCLC.

    We express sincere gratitude to all the authors involved in this study.

    The authors declare that they have no conflict of interest.



    [1] A. Wibisono, A. Wilson, M. Jordan, A variational perspective on accelerated methods in optimization, Proc. Natl. Acad. Sci., 113 (2016), E7351–E7358. https://doi.org/10.1073/pnas.1614734113 doi: 10.1073/pnas.1614734113
    [2] V. Duruisseaux, M. Leok, A variational formulation of accelerated optimization on Riemannian manifolds, SIAM J. Math. Data Sci., 4 (2022), 649–674. https://doi.org/10.1137/21M1395648 doi: 10.1137/21M1395648
    [3] V. Duruisseaux, J. Schmitt, M. Leok, Adaptive Hamiltonian variational integrators and applications to symplectic accelerated optimization, SIAM J. Sci. Comput., 43 (2021), A2949–A2980. https://doi.org/10.1137/20M1383835 doi: 10.1137/20M1383835
    [4] Y. Nesterov, A method of solving a convex programming problem with convergence rate O(1/k2). Sov. Math. Dokl., 27 (1983), 372–376.
    [5] Y. Nesterov, Introductory Lectures on Convex Optimization: A Basic Course, volume 87 of Applied Optimization, Kluwer Academic Publishers, Boston, MA, 2004.
    [6] W. Su, S. Boyd, E. Candes, A differential equation for modeling Nesterov's Accelerated Gradient method: theory and insights. J. Mach. Learn. Res., 17 (2016), 1–43.
    [7] M. Betancourt, M. I. Jordan, A. Wilson, On symplectic optimization. 2018. https://doi.org/10.48550/arXiv.1802.03653
    [8] C. M. Campos, A. Mahillo, D. Martín de Diego, A discrete variational derivation of accelerated methods in optimization. 2021. https://doi.org/10.48550/arXiv.2106.02700
    [9] G. França, M. I. Jordan, R. Vidal, On dissipative symplectic integration with applications to gradient-based optimization. J. Stat. Mech., 2021 (2021), 043402. https://doi.org/10.1088/1742-5468/abf5d4 doi: 10.1088/1742-5468/abf5d4
    [10] G. França, A. Barp, M. Girolami, M. I. Jordan, Optimization on manifolds: A symplectic approach, 07 2021. https://doi.org/10.48550/arXiv.2107.11231
    [11] M. I. Jordan, Dynamical, symplectic and stochastic perspectives on gradient-based optimization, In Proceedings of the International Congress of Mathematicians (ICM 2018), 523–549. https://doi.org/10.1142/9789813272880_0022
    [12] E. Hairer, C. Lubich, G. Wanner, Geometric Numerical Integration, volume 31 of Springer Series in Computational Mathematics, Springer-Verlag, Berlin, second edition, 2006.
    [13] J. P. Calvo, J. M. Sanz-Serna, The development of variable-step symplectic integrators, with application to the two-body problem, SIAM J. Sci. Comp., 14 (1993), 936–952. https://doi.org/10.1137/0914057 doi: 10.1137/0914057
    [14] B. Gladman, M. Duncan, J. Candy, Symplectic integrators for long-time integrations in celestial mechanics. Celestial Mech. Dyn. Astr., 52 (1991), 221–240. https://doi.org/10.1007/BF00048485 doi: 10.1007/BF00048485
    [15] E. Hairer, Variable time step integration with symplectic methods, Appl. Numer. Math., 25 (1997), 219–227. https://doi.org/10.1016/S0168-9274(97)00061-5 doi: 10.1016/S0168-9274(97)00061-5
    [16] K. Zare, V. G. Szebehely, Time transformations in the extended phase-space. Celestial Mech., 11 (1975), 469–482. https://doi.org/10.1007/BF01650285 doi: 10.1007/BF01650285
    [17] J. Hall, M. Leok, Spectral Variational Integrators, Numer. Math., 130 (2015), 681–740. https://doi.org/10.1007/s00211-014-0679-0 doi: 10.1007/s00211-014-0679-0
    [18] J. E. Marsden, M. West, Discrete mechanics and variational integrators. Acta Numer., 10 (2001), 357–514. https://doi.org/10.1017/S096249290100006X doi: 10.1017/S096249290100006X
    [19] T. Lee, M. Tao, M. Leok, Variational symplectic accelerated optimization on Lie groups. 2021.
    [20] M. Tao, T. Ohsawa, Variational optimization on Lie groups, with examples of leading (generalized) eigenvalue problems. In Proceedings of the 23rd International AISTATS Conference, volume 108 of PMLR, 2020.
    [21] F. Alimisis, A. Orvieto, G. Bécigneul, A. Lucchi, A continuous-time perspective for modeling acceleration in Riemannian optimization. In Proceedings of the 23rd International AISTATS Conference, volume 108 of PMLR, 1297–1307, 2020.
    [22] V. Duruisseaux, M. Leok, Accelerated optimization on Riemannian manifolds via discrete constrained variational integrators. J. Nonlinear Sci., 32 (2022). https://doi.org/10.1007/s00332-022-09795-9 doi: 10.1007/s00332-022-09795-9
    [23] V. Duruisseaux, M. Leok, Accelerated optimization on Riemannian manifolds via projected variational integrators. 2022. https://doi.org/10.48550/arXiv.2201.02904
    [24] J. Nash, The imbedding problem for Riemannian manifolds, Ann. Math., 63 (1956), 20–63. https://doi.org/10.2307/1969989 doi: 10.2307/1969989
    [25] H. Whitney, The self-intersections of a smooth n-manifold in 2n-space, Ann. Math., 45 (1944), 220–246. https://doi.org/10.2307/1969265 doi: 10.2307/1969265
    [26] H. Whitney, The singularities of a smooth n-manifold in (2n1)-space, Ann. Math., 45 (1944), 247–293. https://doi.org/10.2307/1969266 doi: 10.2307/1969266
    [27] N. Bou-Rabee, J. E. Marsden, Hamilton–Pontryagin integrators on Lie groups part Ⅰ: Introduction and structure-preserving properties, Found. Comput. Math., 9 (2009), 197–219. https://doi.org/10.1007/s10208-008-9030-4 doi: 10.1007/s10208-008-9030-4
    [28] I. I. Hussein, M. Leok, A. K. Sanyal, A. M. Bloch, A discrete variational integrator for optimal control problems on SO(3). Proceedings of the 45th IEEE Conference on Decision and Control, 6636–6641, 2006. https://doi.org/10.1109/CDC.2006.377818 doi: 10.1109/CDC.2006.377818
    [29] T. Lee, Computational geometric mechanics and control of rigid bodies. Ph.D. dissertation, University of Michigan, 2008.
    [30] T. Lee, M. Leok, N. H. McClamroch, Lie group variational integrators for the full body problem, Comput. Methods Appl. Mech. Engrg., 196 (2007), 2907–2924. https://doi.org/10.1016/j.cma.2007.01.017 doi: 10.1016/j.cma.2007.01.017
    [31] T. Lee, M. Leok, N. H. McClamroch, Lie group variational integrators for the full body problem in orbital mechanics, Celestial Mech. Dyn. Astr., 98 (2007), 121–144. https://doi.org/10.1007/s10569-007-9073-x doi: 10.1007/s10569-007-9073-x
    [32] T. Lee, N. H. McClamroch, M. Leok, A Lie group variational integrator for the attitude dynamics of a rigid body with applications to the 3d pendulum, Proceedings of 2005 IEEE Conference on Control Applications. (CCA 2005), 962–967. https://doi.org/10.1109/CCA.2005.1507254 doi: 10.1109/CCA.2005.1507254
    [33] M. Leok, Generalized Galerkin variational integrators: Lie group, multiscale, and pseudospectral methods, (preprint, arXiv: math.NA/0508360), 2004.
    [34] N. Nordkvist, A. K. Sanyal, A Lie group variational integrator for rigid body motion in SE(3) with applications to underwater vehicle dynamics, In 49th IEEE Conference on Decision and Control (CDC), 5414–5419, 2010. https://doi.org/10.1109/CDC.2010.5717622
    [35] T. Lee, M. Leok, N. H. McClamroch, Lagrangian mechanics and variational integrators on two-spheres, Int. J. Numer. Methods Eng., 79 (2009), 1147–1174. https://doi.org/10.1002/nme.2603 doi: 10.1002/nme.2603
    [36] J. E. Marsden, T. S. Ratiu, Introduction to mechanics and symmetry, volume 17 of Texts in Applied Mathematics. Springer-Verlag, New York, second edition, 1999.
    [37] S. Lall, M. West, Discrete variational Hamiltonian mechanics, J. Phys. A, 39 (2006), 5509–5519. https://doi.org/10.1088/0305-4470/39/19/S11 doi: 10.1088/0305-4470/39/19/S11
    [38] M. Leok, J. Zhang, Discrete Hamiltonian variational integrators, IMA J. Numer. Anal., 31 (2011), 1497–1532. https://doi.org/10.1093/imanum/drq027 doi: 10.1093/imanum/drq027
    [39] J. M. Schmitt, M. Leok, Properties of Hamiltonian variational integrators, IMA J. Numer. Anal., 38 (2017), 377–398. https://doi.org/10.1093/imanum/drx010 doi: 10.1093/imanum/drx010
    [40] M. Leok, T. Shingel, Prolongation-collocation variational integrators, IMA J. Numer. Anal., 32 (2012), 1194–1216. https://doi.org/10.1093/imanum/drr042 doi: 10.1093/imanum/drr042
    [41] J. M. Schmitt, T. Shingel, M. Leok, Lagrangian and Hamiltonian Taylor variational integrators, BIT Numer. Math., 58 (2018), 457–488. https://doi.org/10.1007/s10543-017-0690-9 doi: 10.1007/s10543-017-0690-9
    [42] M. Leok, T. Ohsawa, Variational and geometric structures of discrete Dirac mechanics, Found. Comput. Math., 11 (2011), 529–562. https://doi.org/10.1007/s10208-011-9096-2 doi: 10.1007/s10208-011-9096-2
    [43] H. Yoshimura, J. E. Marsden, Dirac structures in Lagrangian mechanics Part Ⅰ: Implicit Lagrangian systems, J. Geom. Phys., 57 (2006), 133–156. https://doi.org/10.1016/j.geomphys.2006.02.009 doi: 10.1016/j.geomphys.2006.02.009
    [44] H. Yoshimura, J. E. Marsden, Dirac structures in Lagrangian mechanics Part Ⅱ: Variational structures, J. Geom. Phys., 57 (2006), 209–250. https://doi.org/10.1016/j.geomphys.2006.02.012 doi: 10.1016/j.geomphys.2006.02.012
    [45] D. Kingma, J. Ba, Adam: A method for stochastic optimization, International Conference on Learning Representations, 12 2014. https://doi.org/10.48550/arXiv.1412.6980
    [46] J. Duchi, E. Hazan, Y. Singer, Adaptive subgradient methods for online learning and stochastic optimization, J. Mach. Learn. Res., 12 (2011), 2121–2159.
    [47] T. Tieleman, G. Hinton, Coursera: Neural Networks for Machine Learning - Lecture 6.5: RMSprop, 2012.
    [48] P. A. Absil, R. Mahony, R. Sepulchre, Optimization Algorithms on Matrix Manifolds, Princeton University Press, Princeton, NJ, 2008. https://doi.org/10.1515/9781400830244
    [49] N. Boumal, An introduction to optimization on smooth manifolds, 2020.
    [50] J. M. Lee, Introduction to Riemannian Manifolds, volume 176 of Graduate Texts in Mathematics, Springer, Cham, second edition, 2018.
    [51] J. Jost, Riemannian Geometry and Geometric Analysis, Universitext. Springer, Cham, 7th edition, 2017.
    [52] T. Lee, M. Leok, N. H. McClamroch, Global Formulations of Lagrangian and Hamiltonian Dynamics on Manifolds: A Geometric Approach to Modeling and Analysis, Interaction of Mechanics and Mathematics. Springer International Publishing, 2017.
    [53] J. Gallier, J. Quaintance, Differential Geometry and Lie Groups: A Computational Perspective, Geometry and Computing. Springer International Publishing, 2020.
    [54] E. Celledoni, H. Marthinsen, B. Owren, An introduction to Lie group integrators – basics, new developments and applications, J. Comput. Phys., 257 (2014), 1040–1061. https://doi.org/10.1016/j.jcp.2012.12.031 doi: 10.1016/j.jcp.2012.12.031
    [55] E. Celledoni, E. Çokaj, A. Leone, D. Murari, B. Owren, Lie group integrators for mechanical systems, Int. J. Comput. Math., 99 (2022), 58–88. https://doi.org/10.1080/00207160.2021.1966772 doi: 10.1080/00207160.2021.1966772
    [56] S. H. Christiansen, H. Z. Munthe-Kaas, B. Owren, Topics in structure-preserving discretization, Acta Numer., 20 (2011), 1–119. https://doi.org/10.1017/S096249291100002X doi: 10.1017/S096249291100002X
    [57] A. Iserles, H. Z. Munthe-Kaas, S. P. Nørsett, A. Zanna, Lie group methods, Acta Numer., 9 (2000), 215–365. https://doi.org/10.1017/S0962492900002154
    [58] P. E. Crouch, R. L. Grossman, Numerical integration of ordinary differential equations on manifolds, J. Nonlinear Sci., 3 (1993), 1–33. https://doi.org/10.1007/BF02429858 doi: 10.1007/BF02429858
    [59] D. Lewis, J. C. Simo, Conserving algorithms for the dynamics of Hamiltonian systems on Lie groups, J. Nonlinear Sci., 4 (1994), 253–299. https://doi.org/10.1007/BF02430634 doi: 10.1007/BF02430634
    [60] F. Casas, B. Owren, Cost efficient Lie group integrators in the RKMK class, BIT, 43 (2003), 723–742. https://doi.org/10.1023/B:BITN.0000009959.29287.d4 doi: 10.1023/B:BITN.0000009959.29287.d4
    [61] H. Z. Munthe-Kaas, Lie-Butcher theory for Runge-Kutta methods, BIT, 35 (1995), 572–587. https://doi.org/10.1007/BF01739828 doi: 10.1007/BF01739828
    [62] H. Z. Munthe-Kaas, Runge-Kutta methods on Lie groups, Bit Numer. Math., 38 (1998), 92–111. https://doi.org/10.1007/BF02510919 doi: 10.1007/BF02510919
    [63] H. Z. Munthe-Kaas, High order Runge-Kutta methods on manifolds, Appl. Numer. Math., 29 (1999), 115–127. https://doi.org/10.1016/S0168-9274(98)00030-0 doi: 10.1016/S0168-9274(98)00030-0
    [64] S. Blanes, F. Casas, J. A. Oteo, J. Ros, The Magnus expansion and some of its applications, Phys. Rep., 470 (2008), 151–238. https://doi.org/10.1016/j.physrep.2008.11.001 doi: 10.1016/j.physrep.2008.11.001
    [65] A. Iserles, S. P. Nørsett, On the solution of linear differential equations in Lie groups, Phil. Trans. R. Soc. A, 357 (1999), 983–1019. https://doi.org/10.1098/rsta.1999.0362 doi: 10.1098/rsta.1999.0362
    [66] A. Zanna, Collocation and relaxed collocation for the Fer and the Magnus expansions, SIAM J. Numer. Anal., 36 (1999), 1145–1182. https://doi.org/10.1137/S0036142997326616 doi: 10.1137/S0036142997326616
    [67] E. Celledoni, A. Marthinsen, B. Owren, Commutator-free Lie group methods, Future Gener. Comput. Syst., 19 (2003), 341–352. https://doi.org/10.1016/S0167-739X(02)00161-9 doi: 10.1016/S0167-739X(02)00161-9
    [68] G. Wahba, A Least Squares Estimate of Satellite Attitude, SIAM Rev., 7 (1965), 409. https://doi.org/10.1137/1007077 doi: 10.1137/1007077
    [69] V. Duruisseaux, M. Leok, Practical perspectives on symplectic accelerated optimization, 2022. https://doi.org/10.48550/arXiv.2207.11460
  • jgm-15-01-010-s001.pdf
  • This article has been cited by:

    1. Guocan Yu, Yanqin Shen, Xudong Xu, Fangming Zhong, Sherief Ghozy, Anlotinib for refractory advanced non-small-cell lung cancer: A systematic review and meta-analysis, 2020, 15, 1932-6203, e0242982, 10.1371/journal.pone.0242982
    2. Guocan Yu, Qingshan Cai, Xudong Xu, Yanqin Shen, Kan Xu, Sherief Ghozy, Anlotinib-containing regimen for advanced small-cell lung cancer: A protocol of meta-analysis, 2021, 16, 1932-6203, e0247494, 10.1371/journal.pone.0247494
    3. Qiang Li, You-Seok Kim, Ju-Hyun An, Jin-Ah Kwon, Sang-Hyun Han, Woo-Jin Song, Hwa-Young Youn, Anti-tumor effects of rivoceranib against canine melanoma and mammary gland tumour in vitro and in vivo mouse xenograft models, 2021, 17, 1746-6148, 10.1186/s12917-021-03026-1
    4. Zhe Li, Zhibao Liu, Yuanyuan Wu, Huarui Li, Zhen Sun, Chenggang Han, Xiaoling Zhang, Jinghua Zhang, Efficacy and safety of apatinib alone or apatinib plus paclitaxel/docetaxel versus paclitaxel/docetaxel in the treatment of advanced non‐small cell lung cancer: A meta‐analysis, 2021, 12, 1759-7706, 2838, 10.1111/1759-7714.14131
    5. Ningjie Li, Ping Yang, Jun Fang, Transarterial chemoembolization (TACE) plus apatinib vs. TACE alone for hepatocellular carcinoma, 2022, 46, 22107401, 102022, 10.1016/j.clinre.2022.102022
    6. Ling Zhou, Wenchao Zhang, Yi Xiang, Zijun Qian, Jianping Zhou, Lei Ni, Yun Feng, Beili Gao, The apatinib and pemetrexed combination has antitumor and antiangiogenic effects against NSCLC, 2023, 18, 2391-5412, 10.1515/biol-2022-0533
    7. Xin Zhang, Haoran Huo, Yanan Nie, Jiadong Xue, Zengjiang Yuan, Zhenyi Zhang, Apatinib as a Third-Line Treatment for HER2-Positive Metastatic Gastric Cancer: A Multi-Center Single-Arm Cohort Study, 2022, 22, 2093-582X, 408, 10.5230/jgc.2022.22.e33
    8. Haiming Chen, Xinfu Liu, Minhong He, Weiguang Gu, Yiyu Lu, Apatinib successfully administered in malignant tumour patients with severe hypotension: a case series, literature review, and considerations for treatment, 2023, 51, 0300-0605, 10.1177/03000605231177186
    9. Jianping Xu, Xiaoyan Liu, Sheng Yang, Yuankai Shi, Efficacy, safety, and prognostic factors of apatinib plus platinum doublet chemotherapy in advanced non-small cell lung cancer, 2022, 18, 0973-1482, 1425, 10.4103/jcrt.jcrt_1853_21
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1789) PDF downloads(43) Cited by(6)

Figures and Tables

Figures(4)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog