Sonu Kurmi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Sonu Kurmi has a total of 2 co-authors. Here are the top co-authors by number of shared publications:
- Usha Chouhan
Maulana Azad National Institute of Technology
3 shared publications - Abhinav Gupta
Maulana Azad National Institute of Technology (MANIT)
1 shared publication
Latest Publications
- 17 November 2026
- Published by IOP Publishing in Physica Scripta
- Vol. 101 (46) , 465204
AlphaNet7: a multi-head physics-informed neural network for latent parameter discovery in variable-order fractional systems
Solving high-dimensional inverse problems in variable-order fractional dynamical systems from sparse observations remains a major challenge in scientific machine learning. Standard physics-informed neural networks (PINNs) rely on soft residual penalties, which frequently induce gradient pathologies or encounter conservation discrepancies when applied to non-local fractional operators. To address this limitation, we propose AlphaNet7, a constrained multi-head PINN architecture for the simultaneous inverse identification of compartment-specific, time-varying fractional orders and transmission dynamics. The proposed architecture combines a shared feature-extraction backbone with eight independent prediction heads, jointly inferring multiple latent quantities while implicitly integrating the governing phenomenological equations through a custom autograd-compatible piecewise differentiable explicit fractional forward solver. AlphaNet7 is applied to a seven-compartment nonlinear epidemiological system governed by the non-singular Caputo–Fabrizio fractional operator. From two macroscopic surveillance streams, the network estimates a regularization-guided, constraint-dependent set of seven phenomenological effective fractional orders rather than genuine biological memory parameters. In the reference run, inferred compartment orders span α ≈ 0.688 – 0.864 (mean α ― ≈ 0.753 ), whereas a single-head global-order model achieves identical reconstruction error ( RMSE = 0.0028 ) and removing the diversity penalty contracts compartment means to between 0.740 and 0.774 ( RMSE = 0.0021 ). Under regularization-guided optimization, the multi-head model captures non-linear transitions in exposure and recovery alongside quasi-stationary institutional plateaus; this alignment reflects post hoc temporal coincidence with policy shifts rather than causality. Furthermore, the inferred regime yields a modest reduction of roughly 2 % – 3 % ( 1.38 → 1.35 ) in the definition-driven, phenomenological fractional reproduction metric. Because the explicit solver numerically constrains trajectories to the discretized governing dynamics, AlphaNet7 outperforms fixed-order baselines in calibration and out-of-sample forecasting, reducing forecast error by 30.2 % – 63.1 % across rolling hold-out splits under unified normalization. This provides a prior-constrained framework for effective parameter estimation in variable-order fractional systems.
- 17 November 2026
- Published by IOP Publishing in Physica Scripta
- Vol. 101 (46) , 465204
AlphaNet7: a multi-head physics-informed neural network for latent parameter discovery in variable-order fractional systems
Solving high-dimensional inverse problems in variable-order fractional dynamical systems from sparse observations remains a major challenge in scientific machine learning. Standard physics-informed neural networks (PINNs) rely on soft residual penalties, which frequently induce gradient pathologies or encounter conservation discrepancies when applied to non-local fractional operators. To address this limitation, we propose AlphaNet7, a constrained multi-head PINN architecture for the simultaneous inverse identification of compartment-specific, time-varying fractional orders and transmission dynamics. The proposed architecture combines a shared feature-extraction backbone with eight independent prediction heads, jointly inferring multiple latent quantities while implicitly integrating the governing phenomenological equations through a custom autograd-compatible piecewise differentiable explicit fractional forward solver. AlphaNet7 is applied to a seven-compartment nonlinear epidemiological system governed by the non-singular Caputo–Fabrizio fractional operator. From two macroscopic surveillance streams, the network estimates a regularization-guided, constraint-dependent set of seven phenomenological effective fractional orders rather than genuine biological memory parameters. In the reference run, inferred compartment orders span α ≈ 0.688 – 0.864 (mean α ― ≈ 0.753 ), whereas a single-head global-order model achieves identical reconstruction error ( RMSE = 0.0028 ) and removing the diversity penalty contracts compartment means to between 0.740 and 0.774 ( RMSE = 0.0021 ). Under regularization-guided optimization, the multi-head model captures non-linear transitions in exposure and recovery alongside quasi-stationary institutional plateaus; this alignment reflects post hoc temporal coincidence with policy shifts rather than causality. Furthermore, the inferred regime yields a modest reduction of roughly 2 % – 3 % ( 1.38 → 1.35 ) in the definition-driven, phenomenological fractional reproduction metric. Because the explicit solver numerically constrains trajectories to the discretized governing dynamics, AlphaNet7 outperforms fixed-order baselines in calibration and out-of-sample forecasting, reducing forecast error by 30.2 % – 63.1 % across rolling hold-out splits under unified normalization. This provides a prior-constrained framework for effective parameter estimation in variable-order fractional systems.
- 25 February 2025
- Published by Springer Nature
- p. 311-329
A Fractional Order Approach to Mathematical Modelling and Simulations of COVID-19 with Vaccination as a Control Parameter
- 25 February 2025
- Published by Springer Nature
- p. 311-329
A Fractional Order Approach to Mathematical Modelling and Simulations of COVID-19 with Vaccination as a Control Parameter
- Open Access
- 13 June 2022
- Published by Springer Nature in Nonlinear Dynamics
- Vol. 109 (3) , 2185-2201
A multicompartment mathematical model to study the dynamic behaviour of COVID-19 using vaccination as control parameter
To analyse novel coronavirus disease (COVID-19) transmission in India, this article provides an extended SEIR multicompartment model using vaccination as a control parameter. The model considers eight classes of infection: susceptible (\(\mathcal {S}\)), vaccinated (\(\mathcal {V}\)), exposed (\(\mathcal {E}\)), asymptomatic infected (\(\mathcal {A}\)), symptomatic infected (\(\mathcal {I}\)), isolated (\(\mathcal {J}\)), hospitalised (\(\mathcal {H}\)), recovered (\(\mathcal {R}\)). To begin, a mathematical study is performed to demonstrate the suggested model’s uniform boundedness, epidemic equilibrium, and basic reproduction number. The findings indicate that if, \(R_{0}1\) the equilibrium is unstable. Secondly, we examine the effect on those who have received vaccinations with what are deemed optimal values. The suggested model is numerically simulated using MATLAB 14.0, and the results confirm the capacity of the proposed model to provide an accurate forecast of the progress of the epidemic in India. Finally, we examine the impact of immunisation on COVID-19 dissemination.- 13 June 2022
- Published by Springer Nature in Nonlinear Dynamics
- Vol. 109 (3) , 2185-2201
Open AccessA multicompartment mathematical model to study the dynamic behaviour of COVID-19 using vaccination as control parameter
To analyse novel coronavirus disease (COVID-19) transmission in India, this article provides an extended SEIR multicompartment model using vaccination as a control parameter. The model considers eight classes of infection: susceptible (\(\mathcal {S}\)), vaccinated (\(\mathcal {V}\)), exposed (\(\mathcal {E}\)), asymptomatic infected (\(\mathcal {A}\)), symptomatic infected (\(\mathcal {I}\)), isolated (\(\mathcal {J}\)), hospitalised (\(\mathcal {H}\)), recovered (\(\mathcal {R}\)). To begin, a mathematical study is performed to demonstrate the suggested model’s uniform boundedness, epidemic equilibrium, and basic reproduction number. The findings indicate that if, \(R_{0}1\) the equilibrium is unstable. Secondly, we examine the effect on those who have received vaccinations with what are deemed optimal values. The suggested model is numerically simulated using MATLAB 14.0, and the results confirm the capacity of the proposed model to provide an accurate forecast of the progress of the epidemic in India. Finally, we examine the impact of immunisation on COVID-19 dissemination.