Machine Learning Control Based on Approximation of Optimal Trajectories

The paper is devoted to an emerging trend in control—a machine learning control. Despite the popularity of idea learning, there are various interpretations this concept, and urgent need for its strict mathematical formalization. An attempt formalize concept presented paper. concepts unknown function, work area, training set introduced, a formulation problem presented. Based on formulation, control considered. One problems general synthesis It implies finding function that depends state object, which ensures achievement goal with optimal value quality criterion from any initial some admissible region. Supervised unsupervised approaches solving based symbolic regression methods As computational example, spacecraft landing surface Moon considered as supervised set.