The Black-Scholes equation was the revolutionary building block for pricing options, which laid ground for and fuelled further research in derivative pricing. The Universal Approximation Theorem proves that Neural Networks can be used to approximate any continuous function.
Leveraging this property of Neural Networks and the ease of computing gradients using the chain rule, in this project, Physics Informed Gated Neural Networks have been used to approximate the solution of the Black Scholes Equations for pricing European Options.
