Add and Thin: Diffusion for Temporal Point Processes

arXiv:2311.01139v2 Announce Type: replace
Abstract: Autoregressive neural networks within the temporal point process (TPP) framework have become the standard for modeling continuous-time event data. Even though these models can expressively capture event sequences in a one-step-ahead fashion, they are inherently limited for long-term forecasting applications due to the accumulation of errors caused by their sequential nature. To overcome these limitations, we derive ADD-THIN, a principled probabilistic denoising diffusion model for TPPs that operates on entire event sequences. Unlike existing diffusion …

abstract applications arxiv become continuous cs.lg data diffusion errors event fashion forecasting framework long-term modeling networks neural networks process processes standard stat.ml temporal type