notesum.ai
Published at November 7A Bayesian Mixture Model of Temporal Point Processes with Determinantal Point Process Prior
cs.LG
cs.AI
Released Date: November 7, 2024
Authors: Yiwei Dong1, Shaoxin Ye1, Yuwen Cao1, Qiyu Han1, Hongteng Xu1, Hanfang Yang1
Aff.: 1Renmin University of China

| Model | Layer | Purity | ARI |
| Hawkes | None | 0.702 | 0.648 |
| Diagonal Elements of Infectivity Matrix | 0.655 | 0.572 | |
| Base Intensity | 0.739 | 0.626 | |
| RMTPP | None | 0.750 | 0.679 |
| Time Embedding Layer | 0.747 | 0.664 | |
| Output Layer | 0.753 | 0.708 | |
| THP | None | 0.722 | 0.605 |
| Post-attention Feedforward Layer 1 | 0.740 | 0.630 | |
| Post-attention Feedforward Layer 2 | 0.738 | 0.610 | |
| Post-attention Feedforward Layer 3 | 0.745 | 0.647 | |
| Post-attention Feedforward Layer 4 | 0.748 | 0.647 | |
| Output Layer | 0.749 | 0.652 |