Research map / Time series / Causal and generative time series
C-NTPP: Learning Cluster-Aware Neural Temporal 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 |
|---|---|---|---|
| Add event-clustered attention to the global-attention encoder (ablation) | c-NTPP · SYN, FIN, MMC, RET, SO | negative log-likelihood per event: SYN: 2.610 -> 2.557; FIN: 1.195 -> 1.180; MMC: 2.665 -> 2.178; RET: 7.379 -> 7.234; SO: 2.571 -> 2.551 | Helped |
| Add event-clustered attention (ablation) | c-NTPP · RET | RMSE: 91.86 -> 92.42 (mean); 91.86 -> 92.71 (attention) | Hurt |
| Add event-clustered attention (ablation) | c-NTPP · SYN, FIN, MMC, RET, SO | next event marker accuracy: SYN: 75.26 -> 75.26; FIN: 62.63 -> 62.61; MMC: 59.88 -> 69.18; RET: 58.59 -> 59.13; SO: 43.23 -> 43.51 | Mixed |
| Use mean rather than attention aggregation across inferred clusters (ablation) | c-NTPP · SYN, FIN, MMC, RET, SO | negative log-likelihood per event: 2.558 vs 2.557; 1.173 vs 1.180; 2.138 vs 2.178; 7.400 vs 7.234; 2.551 vs 2.692 | Mixed |
| Use mean rather than attention aggregation across inferred clusters (ablation) | c-NTPP · SYN, FIN, MMC, RET, SO | RMSE: 1.516 vs 1.513; 11.35 vs 11.35; 5.506 vs 5.503; 92.42 vs 92.71; 8.661 vs 8.541 | Mixed |
| Use mean rather than attention aggregation across inferred clusters (ablation) | c-NTPP · SYN, FIN, MMC, RET, SO | next event marker accuracy: 75.26 vs 75.26; 62.59 vs 62.61; 73.25 vs 69.18; 58.56 vs 59.13; 43.86 vs 43.51 | Mixed |
| Set maximum cluster number K to 1, making the model degenerate to THP (ablation) | c-NTPP · TPP modeling | negative log-likelihood per event | No clear effect |
| Use a sequential variational cluster model for event-cluster inference | c-NTPP · SEP | NMI: 0.8044 -> 0.9837 | Helped |
| Use a sequential variational cluster model for event-cluster inference | c-NTPP · OVL | NMI: 0.1392 -> 0.4231 | Helped |
Closest papers by meaning
- A Variational Autoencoder for Neural Temporal Point Processes with Dynamic Latent Graphs (AAAI 2024 · similarity 0.89)
- EasyTPP: Towards Open Benchmarking Temporal Point Processes (ICLR 2024 · similarity 0.86)
- Interacting Diffusion Processes for Event Sequence Forecasting (ICML 2024 · similarity 0.86)
- Deep Continuous-Time State-Space Models for Marked Event Sequences (NeurIPS 2025 · similarity 0.85)
- Residual TPP: A Unified Lightweight Approach for Event Stream Data Analysis (ICML 2025 · similarity 0.85)
- A Plug-and-Play Bregman ADMM Module for Inferring Event Branches in Temporal Point Processes (AAAI 2025 · similarity 0.85)
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.