A Physics-Informed Neural Network (PINN) is a type of neural network that incorporates physical laws, typically expressed as differential equations (ODEs and PDEs), into its training process.
Instead of learning solely from data, it is trained to satisfy both the observed data and the governing equations, making it especially useful when data is scarce but physical models are known.
This allows PINNs to solve forward and inverse problems in physics, engineering, and applied mathematics with high flexibility and generalization.