Latent Ordinary Differential Equations for Irregularly-Sampled Time Series

Time series with non-uniform intervals occur in many applications, and are difficult to model using standard recurrent neural networks (RNNs).