The paper proposes a new physics-informed, weighted length-biased Ramos–Louzada (WLBRL) system for survival and reliability analysis, combining statistical inference with physical constraints. The suggested WLBRL distribution explicitly corrects for length-biased sampling, and physics-informed neural networks (PINNs) impose relationships on survival hazards in estimation that are implicitly defined by differential equations. The physics-informed maximum likelihood estimation approach is presented to stabilize parameter inference under noisy, limited data. Vast simulation literature and real-world medical and engineering data have shown that the proposed framework reduces mean squared error by 40% compared to the classical Ramos–Louzada (RL), Weibull, and exponential models. These findings verify that integrating physical laws into weighted survival models enhances estimation accuracy, robustness, and explainability, indicating that the weighted length-biased Ramos–Louzada physics-informed neural network (WLBRL-PINN) system is generally applicable to uncertainty-aware risk assessment in biomedical and engineering systems.
Citation: Mashail M. AL Sobhi. Physics-informed weighted Ramos–Louzada model: integrating statistical and physical constraints for survival analysis[J]. AIMS Mathematics, 2026, 11(3): 6777-6803. doi: 10.3934/math.2026280
The paper proposes a new physics-informed, weighted length-biased Ramos–Louzada (WLBRL) system for survival and reliability analysis, combining statistical inference with physical constraints. The suggested WLBRL distribution explicitly corrects for length-biased sampling, and physics-informed neural networks (PINNs) impose relationships on survival hazards in estimation that are implicitly defined by differential equations. The physics-informed maximum likelihood estimation approach is presented to stabilize parameter inference under noisy, limited data. Vast simulation literature and real-world medical and engineering data have shown that the proposed framework reduces mean squared error by 40% compared to the classical Ramos–Louzada (RL), Weibull, and exponential models. These findings verify that integrating physical laws into weighted survival models enhances estimation accuracy, robustness, and explainability, indicating that the weighted length-biased Ramos–Louzada physics-informed neural network (WLBRL-PINN) system is generally applicable to uncertainty-aware risk assessment in biomedical and engineering systems.
| [1] |
S. M. Ahad, S. P. Ahmad, Characterization and estimation of the length biased Nakagami distribution, Pakistan J. Stat. Oper. Res., 14 (2018), 697–715. https://doi.org/10.18187/pjsor.v14i3.1930 doi: 10.18187/pjsor.v14i3.1930
|
| [2] |
A. Ahmad, S. P. Ahmad, A. Ahmed, Length-biased weighted Lomax distribution: statistical properties and application, Pakistan J. Stat. Oper. Res., 12 (2016), 245–255. https://doi.org/10.18187/pjsor.v12i2.1178 doi: 10.18187/pjsor.v12i2.1178
|
| [3] |
R. S. Al-Saffar, M. W. Neamah, On a new double weighted exponential-pareto distribution: properties and estimation, Periodicals Eng. Nat. Sci., 10 (2022), 330–342. https://doi.org/10.21533/pen.v10i1.2614 doi: 10.21533/pen.v10i1.2614
|
| [4] |
R. G. Cox, The deformation of a drop in a general time-dependent fluid flow, J. Fluid Mech., 37 (1969), 601–623. https://doi.org/10.1017/S0022112069000759 doi: 10.1017/S0022112069000759
|
| [5] | R. A. Ganaie, V. Rajagopalan, A. A. Rather, On weighted two parameter quasi Shanker distribution with properties and its applications, Int. J. Stat. Reliab. Eng., 7 (2020), 1–12. |
| [6] |
A. Hassan, M. A. Dar, B. A. Peer, B. A. Para, A new generalization of Pranav distribution using weighted technique, Int. J. Sci. Res. Math. Stat. Sci., 6 (2019), 25–32. https://doi.org/10.26438/ijsrmss/v6i1.2532 doi: 10.26438/ijsrmss/v6i1.2532
|
| [7] |
A. Hassan, S. A. Wani, B. A. Para, On three parameter weighted quasi Lindley distribution: properties and applications, Int. J. Sci. Res. Math. Stat. Sci., 5 (2018), 210–224. https://doi.org/10.26438/ijsrmss/v5i5.210224 doi: 10.26438/ijsrmss/v5i5.210224
|
| [8] |
M. Mohiuddin, S. A Dar, A. A Khan, M. A. Ahajeeth, On weighted Nwikpe distribution: properties and applications, Inf. Sci. Lett., 11 (2022), 14. https://doi.org/10.18576/isl/110110 doi: 10.18576/isl/110110
|
| [9] |
G. P. Patil, C. R. Rao, Weighted distributions and size-biased sampling with applications to wildlife populations and human families, Biometrics, 34 (1978), 179–189. https://doi.org/10.2307/2530008 doi: 10.2307/2530008
|
| [10] |
B. A. Para, T. R. Jan, On three parameter weighted pareto type Ⅱ distribution: properties and applications in medical sciences, Appl. Math. Inf. Sci. Lett., 6 (2018), 13–26. https://doi.org/10.18576/amisl/060103 doi: 10.18576/amisl/060103
|
| [11] |
A. A. Rather, C. Subramanian, Characterization and estimation of length biased weighted generalized uniform distribution, Int. J. Sci. Res. Math. Stat. Sci., 5 (2018), 72–76. https://doi.org/10.26438/ijsrmss/v5i5.7276 doi: 10.26438/ijsrmss/v5i5.7276
|
| [12] |
A. A. Rather, C. Subramanian, Exponentiated Mukherjee-Islam distribution, J. Stat. Appl. Probab., 7 (2018), 357–361. https://doi.org/10.18576/jsap/070213 doi: 10.18576/jsap/070213
|
| [13] |
H. M. Reyad, S. A. Othman, A. A. Moussa, The length-biased weighted Erlang distribution, Asian Res. J. Math., 6 (2017), 1–15. https://doi.org/10.9734/ARJOM/2017/35963 doi: 10.9734/ARJOM/2017/35963
|
| [14] |
P. C. V. Deusen, Fitting assumed distributions to horizontal point sample diameters, For. Sci., 32 (1986), 146–148. https://doi.org/10.1093/forestscience/32.1.146 doi: 10.1093/forestscience/32.1.146
|
| [15] | W. G. Warren, Statistical distributions in forestry and forest products research, In: G. P. Patil, S. Kotz, J. K. Ord, A modern course on statistical distributions in scientific work, Springer, 1975,369–384. https://doi.org/10.1007/978-94-010-1845-6_27 |
| [16] |
A. Bihlo, Improving physics-informed neural networks with meta-learned optimization, J. Mach. Learn. Res., 25 (2024), 1–26. https://doi.org/10.5555/3722577.3722591 doi: 10.5555/3722577.3722591
|
| [17] |
P. Rathore, W. Lei, Z. Frangella, L. Lu, M. Udell, Challenges in training PINNs: a loss landscape perspective, arXiv, 2024. https://doi.org/10.48550/arXiv.2402.01868 doi: 10.48550/arXiv.2402.01868
|
| [18] |
R. Brociek, M. Pleszczyński, D. A. Mughal, On the performance of physics-based neural networks for symmetric and asymmetric domains: a comparative study and hyperparameter analysis, Symmetry, 17 (2025), 1698. https://doi.org/10.3390/sym17101698 doi: 10.3390/sym17101698
|
| [19] |
K. Bobzin, H. Heinemann, A. Dokhanchi, Physics-informed neural networks for predicting particle properties in plasma spraying, J. Therm. Spray Technol., 34 (2025), 885–892. https://doi.org/10.1007/s11666-025-01965-x doi: 10.1007/s11666-025-01965-x
|
| [20] |
D. Qayoom, A. A. Rather, M. Alam, Length-biased Burhan distribution: properties, simulation, and applications, Lobachevskii J. Math., 46 (2025), 1374–1390. https://doi.org/10.1134/S1995080225605387 doi: 10.1134/S1995080225605387
|
| [21] |
A. A. Rather, G. Ö. Kadilar, The weighted power Lindley distribution with applications on lifetime data, Pakistan J. Stat. Oper. Res., 16 (2020), 225–237. https://doi.org/10.18187/pjsor.v16i2.2931 doi: 10.18187/pjsor.v16i2.2931
|
| [22] | V. Rajagopalan, R. A. Ganaie, A. A. Rather, A new length biased distribution with applications, Sci. Technol. Dev., 8 (2019), 161–174. |
| [23] |
Y. Martinez, L. Rojas, A. Peña, M. Valenzuela, J. Garcia, Physics-informed neural networks for the structural analysis and monitoring of railway bridges: a systematic review, Mathematics, 13 (2025), 1571. https://doi.org/10.3390/math13101571 doi: 10.3390/math13101571
|
| [24] |
P. Feigl, M. Zelen, Estimation of exponential survival probabilities with concomitant information, Biometrics, 21 (1965), 826–838. https://doi.org/10.2307/2528247 doi: 10.2307/2528247
|
| [25] | S. Sen, I. Ghosh, K. Zografos, A new flexible family of distributions based on the inverse Weibull generator with applications, J. Appl. Stat., 46 (2019), 663–682. |
| [26] |
M. Rasheed, Analyzing applications and properties of the exponential continuous distribution in reliability and survival analysis, J. Positive Sci., 4 (2023), 71–79. https://doi.org/10.52688/ASP30767 doi: 10.52688/ASP30767
|
| [27] |
A. M. Roy, R. Bose, V. Sundararaghavan, R. Arróyave, Deep learning-accelerated computational framework based on physics informed neural network for the solution of linear elasticity, Neural Networks, 162 (2023), 472–489. https://doi.org/10.1016/j.neunet.2023.03.014 doi: 10.1016/j.neunet.2023.03.014
|
| [28] |
A. M. Roy, S. Guha, A data-driven physics-constrained deep learning computational framework for solving von Mises plasticity, Eng. Appl. Artif. Intell., 122 (2023), 106049. https://doi.org/10.1016/j.engappai.2023.106049 doi: 10.1016/j.engappai.2023.106049
|
| [29] |
R. Bose, A. M. Roy, Invariance embedded physics-infused deep neural network-based sub-grid scale models for turbulent flows, Eng. Appl. Artif. Intell., 128 (2024), 107483. https://doi.org/10.1016/j.engappai.2023.107483 doi: 10.1016/j.engappai.2023.107483
|
| [30] |
A. M. Roy, S. Guha, V. Sundararaghavan, R. Arróyave, Physics-infused deep neural network for solution of non-associative Drucker–Prager elastoplastic constitutive model, J. Mech. Phys. Solids, 185 (2024), 105570. https://doi.org/10.1016/j.jmps.2024.105570 doi: 10.1016/j.jmps.2024.105570
|
Figures(9) / Tables(6)
Mashail M. AL Sobhi. Physics-informed weighted Ramos–Louzada model: integrating statistical and physical constraints for survival analysis[J]. AIMS Mathematics, 2026, 11(3): 6777-6803. doi: 10.3934/math.2026280
DownLoad: