Research map / Probabilistic and scientific modeling / Molecules and proteins
The Neural Hawkes Process: A Neurally Self-Modulating Multivariate Point Process
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace the additive Hawkes model with the neural Hawkes model | Retweets · Retweets | held-out log-likelihood | Helped |
| Add the neural recurrent component to the decomposable self-modulating model | Retweets · Retweets | held-out log-likelihood | Helped |
| Allow negative base rates and interaction weights with a positive transfer function | Hawkes process · MemeTrack | cross-entropy per event: -800 to -15 | Helped |
| Replace decomposable interactions with a neural recurrent state | D-SM-MPP · MemeTrack | cross-entropy per event: -15 to around -8 | Helped |
| Fit the neural Hawkes model rather than Hawkes to censored event streams | Hawkes process · Synthetic missing-data datasets | generalization | Helped |
| Use neural Hawkes rather than the comparison recurrent point-process model | Prediction model · Medical, social, and financial prediction datasets | event type prediction error rate | Helped |
| Use neural Hawkes rather than the comparison recurrent point-process model | Prediction model · Medical, social, and financial prediction datasets | time prediction RMSE | Mixed |
| Increase hidden units from 1 to 256 (ablation) | Neural Hawkes · Retweets | log-likelihood: -6.51 -> -6.10 | Helped |
| Halve the number of hidden units (ablation) | Neural Hawkes · Several datasets | held-out log-likelihood: decreasing held-out log-likelihood by < 0.2% relative | No clear effect |
Closest papers by meaning
- Hyper Hawkes Processes: Interpretable Models of Marked Temporal Point Processes (NeurIPS 2026 · similarity 0.90)
- Transformer Embeddings of Irregularly Spaced Events and Their Participants (ICLR 2022 · similarity 0.90)
- Proper Loss Functions for Nonlinear Hawkes Processes (AAAI 2018 · similarity 0.89)
- Self-Attentive Hawkes Process (ICML 2020 · similarity 0.89)
- Interacting Diffusion Processes for Event Sequence Forecasting (ICML 2024 · similarity 0.86)
- A Variational Autoencoder for Neural Temporal Point Processes with Dynamic Latent Graphs (AAAI 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.