Abstract
The problem of determining the underlying dynamics of a system when onlygiven data of its state over time has challenged scientists for decades. Inthis paper, the approach of using machine learning to model the {\em updates}of the phase space variables is introduced; this is done as a function of thephase space variables. (More generally, the modeling is done over the jet spaceof the variables.) This approach is shown to accurately replicate the dynamicsfor the examples of the harmonic oscillator, the pendulum, and the Duffingoscillator; the underlying differential equation is also accurately recoveredin each example. In addition, the results in no way depend on how the data issampled over time (i.e., regularly or irregularly). It is demonstrated thatthis approach (named "FJet") is similar to the model resulting from a Taylorseries expansion of the Runge-Kutta (RK) numerical integration scheme. Thisanalogy confers the advantage of explicitly revealing the appropriate functionsto use in the modeling, as well as revealing the error estimate for theupdates. Thus, this new approach can be thought of as a way to determine thecoefficients of an RK scheme by machine learning. Finally, it is shown in theundamped harmonic oscillator example that the stability of the updates isstable for $10^9$ times longer than with $4$th-order RK.