Fast and Flexible Temporal Point Processes with Triangular Maps

Neural Information Processing Systems 

Temporal point process (TPP) models combined with recurrent neural networks provide a powerful framework for modeling continuous-time event data.