Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
97 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Unveiling Latent Causal Rules: A Temporal Point Process Approach for Abnormal Event Explanation (2402.05946v2)

Published 3 Feb 2024 in cs.LG and cs.AI

Abstract: In high-stakes systems such as healthcare, it is critical to understand the causal reasons behind unusual events, such as sudden changes in patient's health. Unveiling the causal reasons helps with quick diagnoses and precise treatment planning. In this paper, we propose an automated method for uncovering "if-then" logic rules to explain observational events. We introduce temporal point processes to model the events of interest, and discover the set of latent rules to explain the occurrence of events. To achieve this, we employ an Expectation-Maximization (EM) algorithm. In the E-step, we calculate the likelihood of each event being explained by each discovered rule. In the M-step, we update both the rule set and model parameters to enhance the likelihood function's lower bound. Notably, we optimize the rule set in a differential manner. Our approach demonstrates accurate performance in both discovering rules and identifying root causes. We showcase its promising results using synthetic and real healthcare datasets.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (37)
  1. Fast algorithms for mining association rules. In Proc. 20th int. conf. very large data bases, VLDB, volume 1215, pages 487–499. Santiago, Chile, 1994.
  2. Septic shock. The Lancet, 365(9453):63–78, 2005.
  3. Neclatclosed: A vertical algorithm for mining frequent closed itemsets. Expert Systems with Applications, 174:114738, 2021.
  4. Disorders of sodium and water balance in hospitalized patients. Canadian Journal of Anesthesia/Journal canadien d’anesthésie, 56(2):151–167, 2009.
  5. Diastolic arterial blood pressure: a reliable early predictor of survival in human septic shock. Journal of Trauma and Acute Care Surgery, 64(5):1188–1195, 2008.
  6. Jonathan Crabbé and Mihaela Van Der Schaar. Explaining time series predictions with dynamic masks. In International Conference on Machine Learning, pages 2166–2177. PMLR, 2021.
  7. 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.
  8. The relationship between the arteriovenous carbon dioxide gradient and cardiac index. Journal of critical care, 8(4):217–221, 1993.
  9. Graphical modeling for multivariate hawkes processes with nonparametric link functions. Journal of Time Series Analysis, 38(2):225–242, 2017.
  10. Fast vertical mining of sequential patterns using co-occurrence information. In Advances in Knowledge Discovery and Data Mining: 18th Pacific-Asia Conference, PAKDD 2014, Tainan, Taiwan, May 13-16, 2014. Proceedings, Part I 18, pages 40–52. Springer, 2014a.
  11. Vgen: fast vertical mining of sequential generator patterns. In Data Warehousing and Knowledge Discovery: 16th International Conference, DaWaK 2014, Munich, Germany, September 2-4, 2014. Proceedings 16, pages 476–488. Springer, 2014b.
  12. Erminer: sequential rule mining using equivalence classes. In Advances in Intelligent Data Analysis XIII: 13th International Symposium, IDA 2014, Leuven, Belgium, October 30–November 1, 2014. Proceedings 13, pages 108–119. Springer, 2014c.
  13. Blood pressure and arterial lactate level are early indicators of short-term survival in human septic shock. Intensive care medicine, 22:17–25, 1996.
  14. A unified theory of sepsis-induced acute kidney injury: inflammation, microcirculatory dysfunction, bioenergetics and the tubular cell adaptation to injury. Shock (Augusta, Ga.), 41(1):3, 2014.
  15. Mimic-iv, a freely accessible electronic health record dataset. Scientific Data, 10(1):1–9, 2023.
  16. The artificial intelligence clinician learns optimal treatment strategies for sepsis in intensive care. Nature medicine, 24(11):1716–1720, 2018.
  17. Renal blood flow in experimental septic acute renal failure. Kidney international, 69(11):1996–2002, 2006.
  18. International sepsis definitions conference, 2003. In International conference on machine learning. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference …, 2001.
  19. Temporal logic point processes. In International Conference on Machine Learning, pages 5990–6000. PMLR, 2020.
  20. Explaining point processes by learning interpretable temporal logic rules. In International Conference on Learning Representations, 2021.
  21. Central venous-to-arterial carbon dioxide partial pressure difference in early resuscitation from septic shock: a prospective observational study. European Journal of Anaesthesiology| EJA, 31(7):371–380, 2014.
  22. Venous hypercarbia associated with severe sepsis and systemic hypoperfusion. Critical care medicine, 18(6):585–589, 1990.
  23. The neural hawkes process: A neurally self-modulating multivariate point process. Advances in neural information processing systems, 30, 2017.
  24. Barry A Mizock. Alterations in carbohydrate metabolism during stress: a review of the literature. The American journal of medicine, 98(1):75–84, 1995.
  25. Neural nearest neighbors networks. Advances in Neural information processing systems, 31, 2018.
  26. Cynthia Rudin. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature machine intelligence, 1(5):206–215, 2019.
  27. Suchi Saria. Individualized sepsis treatment using reinforcement learning. Nature medicine, 24(11):1641–1642, 2018.
  28. The third international consensus definitions for sepsis and septic shock (sepsis-3). Jama, 315(8):801–810, 2016.
  29. Ashwin Srinivasan. The aleph manual, 2001.
  30. Modeling the intensity function of point process via recurrent neural networks. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 31, 2017.
  31. Reparameterizable subset sampling via continuous relaxations. arXiv preprint arXiv:1901.10517, 2019.
  32. Learning granger causality for hawkes processes. In International conference on machine learning, pages 1717–1726. PMLR, 2016.
  33. Weighted clock logic point process. In The Eleventh International Conference on Learning Representations, 2023.
  34. Self-attentive hawkes process. In International conference on machine learning, pages 11183–11193. PMLR, 2020.
  35. Learning neural point processes with latent graphs. In Proceedings of the Web Conference 2021, pages 1495–1505, 2021.
  36. Hypocalcemia: a pervasive metabolic abnormality in the critically ill. American journal of kidney diseases, 37(4):689–698, 2001.
  37. Transformer hawkes process. In International conference on machine learning, pages 11692–11702. PMLR, 2020.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (4)
  1. Yiling Kuang (1 paper)
  2. Chao Yang (334 papers)
  3. Yang Yang (884 papers)
  4. Shuang Li (203 papers)
Citations (1)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

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