Neural ODEs are neural network which are based on differential equtaions, where normal networks are x = f(x) on. the otherhand neural ODE are dx/dt = f(x). This makes the neural networks learn the change.
In this repo over a Circuit components dataset the model of Neural ODE is trained and is compared with the preformance of Normal Neural Networks
accuracy of NODE: 98% and accuracy of NN: 97.4%