Perturbed Perceptron's Input to Derive Schrödinger Equation in Artificial Neural Networks

With the arrival of powerful computers and advanced algorithms, the searching of new fundamental equations describing our universe, is expected. In this manner, artificial intelligence might to be able to reproduce known equations in physics, for example. By using known equations, the exploration and potential discover on new mathematical structures that could be hidden, can be successfully done. Because of this, in this paper an engineered perceptron is proposed to derive in a straightforward manner the well-known Schrodinger equation. Assumptions ¨ about quantum mechanics neither related formalism were not used. Instead, perceptron is perturbed in their inputs. Once summation is applied, activation is carried out by setting the nullity of that. Therefore, a mathematical scheme to extract the physics meaning of this artificial neuron, is applied. Along the involved processes the evolution operator emerges inherently, fact that smooth the path to Schrodinger equation