Papers
Topics
Authors
Recent
Search
2000 character limit reached

TransFeat-TPP: An Interpretable Deep Covariate Temporal Point Processes

Published 23 Jul 2024 in cs.LG | (2407.16161v1)

Abstract: The classical temporal point process (TPP) constructs an intensity function by taking the occurrence times into account. Nevertheless, occurrence time may not be the only relevant factor, other contextual data, termed covariates, may also impact the event evolution. Incorporating such covariates into the model is beneficial, while distinguishing their relevance to the event dynamics is of great practical significance. In this work, we propose a Transformer-based covariate temporal point process (TransFeat-TPP) model to improve the interpretability of deep covariate-TPPs while maintaining powerful expressiveness. TransFeat-TPP can effectively model complex relationships between events and covariates, and provide enhanced interpretability by discerning the importance of various covariates. Experimental results on synthetic and real datasets demonstrate improved prediction accuracy and consistently interpretable feature importance when compared to existing deep covariate-TPPs.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (42)
  1. G. Adelfio and M. Chiodi. Including covariates in a space-time point process with application to seismicity. Statistical Methods & Applications, 30:947–971, 2021.
  2. Permutation importance: a corrected feature importance measure. Bioinformatics, 26(10):1340–1347, 2010.
  3. Explaining the road accident risk: Weather effects. Accident Analysis & Prevention, 60:456–465, 2013.
  4. An introduction to the theory of point processes: volume I: elementary theory and methods. Springer, 2003.
  5. Recurrent marked temporal point processes: Embedding event history to vector. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 1555–1564, 2016.
  6. The joint effect of weather and lighting conditions on injury severities of single-vehicle accidents. Analytic methods in accident research, 27:100124, 2020.
  7. Spatial point pattern analysis and its application in geographical epidemiology. Transactions of the Institute of British geographers, pages 256–274, 1996.
  8. A. G. Hawkes. Spectra of some self-exciting and mutually exciting point processes. Biometrika, 58(1):83–90, 1971.
  9. Spatial–seasonal characteristics and critical impact factors of pm2. 5 concentration in the beijing–tianjin–hebei urban agglomeration. PLoS one, 13(9):e0201364, 2018.
  10. B. Jahnel and W. König. Probabilistic Methods in Telecommunications. Springer, 2020.
  11. Multi-task learning using uncertainty to weigh losses for scene geometry and semantics. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7482–7491, 2018.
  12. J. T. Kent. Information gain and a general measure of correlation. Biometrika, 70(1):163–173, 1983.
  13. J. F. C. Kingman. Poisson processes, volume 3. Clarendon Press, 1992.
  14. Investigating the characteristics and source analyses of pm2. 5 seasonal variations in chengdu, southwest china. Chemosphere, 243:125267, 2020.
  15. Ecological information from spatial patterns of plants: insights from point process theory. Journal of Ecology, 97(4):616–628, 2009.
  16. Temporal logic point processes. In H. D. III and A. Singh, editors, Proceedings of the 37th International Conference on Machine Learning, volume 119 of Proceedings of Machine Learning Research, pages 5990–6000. PMLR, 13–18 Jul 2020. URL https://proceedings.mlr.press/v119/li20p.html.
  17. Explaining point processes by learning interpretable temporal logic rules. In International Conference on Learning Representations, 2022. URL https://openreview.net/forum?id=P07dq7iSAGr.
  18. Traffic accident modelling via self-exciting point processes. Reliability Engineering & System Safety, 180:312–320, 2018.
  19. Analysis of the influence of precipitation and wind on pm2. 5 and pm10 in the atmosphere. Advances in Meteorology, 2020:1–13, 2020.
  20. H. Mei and J. M. Eisner. The neural hawkes process: A neurally self-modulating multivariate point process. Advances in neural information processing systems, 30, 2017.
  21. S. Meyer. Spatio-temporal infectious disease epidemiology based on point processes. PhD thesis, Institut für Statistik, 2010.
  22. Log gaussian cox processes. Scandinavian journal of statistics, 25(3):451–482, 1998.
  23. Y. Ogata. On lewis’ simulation method for point processes. IEEE transactions on information theory, 27(1):23–31, 1981.
  24. Deep mixture point processes: Spatio-temporal event prediction with rich contextual information. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pages 373–383, 2019.
  25. Dynamic hawkes processes for discovering time-evolving communities’ states behind diffusion processes. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 1276–1286, 2021.
  26. Fully neural network based model for general temporal point processes. Advances in neural information processing systems, 32, 2019.
  27. L. Paninski. Maximum likelihood estimation of cascade point-process neural encoding models. Network: Computation in Neural Systems, 15(4):243–262, 2004.
  28. A. Reinhart and J. Greenhouse. Self-exciting point processes with spatial covariates. Journal of the Royal Statistical Society. Series C (Applied Statistics), 67(5):1305–1329, 2018.
  29. Intensity-free learning of temporal point processes. arXiv preprint arXiv:1909.12127, 2019.
  30. Influence-aware attention for multivariate temporal point processes. In Conference on Causal Learning and Reasoning, pages 499–517. PMLR, 2023.
  31. Feature importance estimation with self-attention networks. arXiv preprint arXiv:2002.04464, 2020.
  32. A deep learning framework for predicting burglaries based on multiple contextual factors. Expert Systems with Applications, 199:117042, 2022.
  33. A point process framework for relating neural spiking activity to spiking history, neural ensemble, and extrinsic covariate effects. Journal of neurophysiology, 93(2):1074–1089, 2005.
  34. Poisson point process models solve the" pseudo-absence problem" for presence-only data in ecology. The Annals of Applied Statistics, pages 1383–1402, 2010.
  35. E. W. Weisstein. Correlation coefficient. https://mathworld. wolfram. com/, 2006.
  36. Modeling the intensity function of point process via recurrent neural networks. In S. Singh and S. Markovitch, editors, Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence, February 4-9, 2017, San Francisco, California, USA, pages 1597–1603. AAAI Press, 2017.
  37. Self-attentive hawkes process. In International conference on machine learning, pages 11183–11193. PMLR, 2020.
  38. Learning neural point processes with latent graphs. In Proceedings of the Web Conference 2021, pages 1495–1505, 2021.
  39. Seasonal trends in pm2. 5 source contributions in beijing, china. Atmospheric Environment, 39(22):3967–3976, 2005.
  40. Efficient inference for dynamic flexible interactions of neural populations. Journal of Machine Learning Research, 23(211):1–49, 2022.
  41. H. Zou. The adaptive lasso and its oracle properties. Journal of the American statistical association, 101(476):1418–1429, 2006.
  42. Transformer hawkes process. In International conference on machine learning, pages 11692–11702. PMLR, 2020.

Summary

No one has generated a summary of this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Sign Up to Summarize

Paper to Video (Beta)

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Sign Up to Generate

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now

Continue Learning

We haven't generated follow-up questions for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now

Collections

Sign up for free to add this paper to one or more collections.

User Circle Single Streamline Icon: https://streamlinehq.com Sign Up