Langevin algorithms for Markovian Neural Networks and Deep Stochastic control

The dataset used in the paper is a stochastic control problem, where the control is parametrized by a neural network calibrated by gradient descent.

BibTex: