Abstract: In this paper, physics informed neural networks are used for numerical approximation of partial differential equations. The data which is used in the process is generated by Latin Hypercube ...