Integrability Ex Machina

Abstract Determining whether a dynamical system is integrable generally difficult task which currently done on case by basis requiring large human input. Here we propose and test an automated method to search for the existence of relevant structures, Lax pair connection respectively. By formulating this as optimization problem, are able identify appropriate structures via machine learning techniques. We our standard systems classical integrability find that can single out some deformations system. Due ambiguity in defining algorithm identifies novel pairs be easily verified analytically.