Learning Point Processes using Recurrent Graph Network
This work addresses the challenge of efficiently modeling event sequences for applications like temporal data analysis, though it appears incremental as it builds on existing point process and graph network methods.
The paper tackles the problem of predicting discrete marked event sequences by learning complex stochastic processes, introducing a Recurrent Graph Network (RGN) that reduces time and space complexity from O(N^2) to O(|Y|^2) and improves performance in log-likelihood, prediction, and goodness-of-fit tasks compared to state-of-the-art Transformer-based architectures.
We present a novel Recurrent Graph Network (RGN) approach for predicting discrete marked event sequences by learning the underlying complex stochastic process. Using the framework of Point Processes, we interpret a marked discrete event sequence as the superposition of different sequences each of a unique type. The nodes of the Graph Network use LSTM to incorporate past information whereas a Graph Attention Network (GAT Network) introduces strong inductive biases to capture the interaction between these different types of events. By changing the self-attention mechanism from attending over past events to attending over event types, we obtain a reduction in time and space complexity from $\mathcal{O}(N^2)$ (total number of events) to $\mathcal{O}(|\mathcal{Y}|^2)$ (number of event types). Experiments show that the proposed approach improves performance in log-likelihood, prediction and goodness-of-fit tasks with lower time and space complexity compared to state-of-the art Transformer based architectures.