pith. sign in

arxiv: 2606.05636 · v1 · pith:FZ3MAHTLnew · submitted 2026-06-04 · 💻 cs.LG

StableRCA: Robust Graph-Agnostic Mechanism-Level Root Cause Analysis

Xiaoyu Lin , Nicholas Tagliapietra , Kehan Li , Lavdim Halilaj , Juergen Luettin This is my paper

Pith reviewed 2026-06-28 02:26 UTC · model grok-4.3

classification 💻 cs.LG
keywords root cause analysisMarkov boundariesconditional distribution shiftsintervention detectiongraph-agnostic methodscausal mechanismssystem diagnostics
0
0 comments X

The pith

Intervention targets can be identified with probability converging exponentially in sample size by estimating local Markov boundaries and detecting conditional distribution shifts within them.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops StableRCA to locate the variables whose mechanisms were altered during abnormal system behavior. It does so locally by recovering Markov boundaries around candidate variables and testing whether their conditional distributions have shifted, without constructing or assuming a global causal graph. Under the conditions that the boundaries are recovered faithfully and the shifts are non-degenerate, the probability of correct identification grows exponentially with the number of samples. This matters for domains such as manufacturing and cloud systems because it removes the need for accurate full-graph estimation while still distinguishing structural causes from mere marginal anomalies.

Core claim

StableRCA identifies intervention targets by estimating local Markov boundaries and detecting conditional distribution shifts within them, leveraging the Independent Causal Mechanism principle. Under faithful Markov boundary recovery and non-degenerate mechanism shifts, the probability of correctly identifying the targets converges exponentially in the sample size. Experiments on synthetic and five real-world datasets show the approach remains effective when graphs are misspecified, when multiple targets are present, and across different application scales.

What carries the argument

Local Markov boundary estimation paired with conditional distribution shift detection inside those boundaries, which isolates intervened mechanisms without global graph construction.

If this is right

  • The method tolerates inaccurate or missing global causal graphs.
  • It continues to work when several variables are intervened on at once.
  • Computation remains feasible as the number of variables grows large.
  • Performance holds across manufacturing, cloud computing, and healthcare data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The local nature of the procedure could be combined with streaming data to monitor live systems without periodic full-graph re-estimation.
  • If boundary recovery is only approximately faithful, the method might still produce useful rankings of candidate causes rather than exact identification.
  • The same local-shift test could be applied inside existing anomaly detection pipelines to move from symptom detection to mechanism-level diagnosis.

Load-bearing premise

The exponential convergence result requires that Markov boundaries are recovered faithfully and that the mechanism shifts are non-degenerate.

What would settle it

A controlled experiment in which Markov boundary recovery is deliberately made unfaithful (for example by adding hidden variables that violate the faithfulness assumption) and the observed identification probability fails to increase exponentially with sample size.

Figures

Figures reproduced from arXiv: 2606.05636 by Juergen Luettin, Kehan Li, Lavdim Halilaj, Nicholas Tagliapietra, Xiaoyu Lin.

Figure 1
Figure 1. Figure 1: Motivation and Contribution. : a) Traditional Graph-Dependent RCA, and b) StableRCA Population-Level Mechanism RCA. formulate RCA as intervention-target identification under Structural Causal Models (SCMs). For example, CIRCA identifies root causes by measuring changes in conditional distributions given parent variables on a causal Bayesian network [Li et al., 2022]. Under the assumption that the SCM is kn… view at source ↗
Figure 2
Figure 2. Figure 2: Illustration of the StableRCA framework. It comprises three main phases: 1) Marginal [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Top-1 accuracy across 5 different real-world [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Causal graph of ProRCA dataset. front-end catalogue carts user orders payment orders-db queue-master rabbitmq shipping user-db session-db catalogue-db carts-db [PITH_FULL_IMAGE:figures/full_fig_p025_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Causal graph of Sockshop dataset. randomly sampled abnormal instance. For graph-based methods, we use the available causal graph whenever possible. Specifically, the ground-truth causal graphs are used for ProRCA, CausalMan, and CausalChambers. For Sock-Shop, we use the edge-reversed service call graph as a proxy causal graph. For RCAEval, where ground-truth causal graphs are unavailable, we estimate a gra… view at source ↗
Figure 6
Figure 6. Figure 6: Causal graph of Causal-Chamber dataset. is treated as the outcome variable. Five types of anomalies are injected into the system: CPU hog (cpu), memory leak (mem), disk I/O stress (disk), network delay (delay), and packet loss (loss). Each anomaly type is independently injected into one of five services — CARTS, CATALOGUE, ORDERS, PAYMENT, and USER. For each anomaly–service pair, five independent replicate… view at source ↗
Figure 7
Figure 7. Figure 7: Case-level performance heatmap of different methods on ProRCA. [PITH_FULL_IMAGE:figures/full_fig_p028_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Case-level performance heatmap of different methods on Sock-shop. [PITH_FULL_IMAGE:figures/full_fig_p028_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Case-level performance heatmap of different methods on Causal-Chamber. [PITH_FULL_IMAGE:figures/full_fig_p029_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Performance heatmap of different methods on CausalMan. [PITH_FULL_IMAGE:figures/full_fig_p030_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Performance heatmap of different methods on RCAEval, grouped by dataset. [PITH_FULL_IMAGE:figures/full_fig_p030_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Performance heatmap of different methods on RCAEval, grouped by fault type. [PITH_FULL_IMAGE:figures/full_fig_p031_12.png] view at source ↗
read the original abstract

Root-Cause Analysis (RCA) seeks to identify the variables responsible for abnormal system behavior in complex domains such as manufacturing, cloud computing, and healthcare. Existing approaches face a critical bottleneck: graph-based causal methods can identify intervention targets but typically require a known or accurately estimated causal graph, while graph-free statistical methods either localize marginal anomalies rather than structural causes, or rely on restrictive assumptions about graph structure or functional form. We propose StableRCA, a local mechanism-level RCA framework that avoids global graph discovery by estimating local Markov boundaries and detecting conditional distribution shifts within them. Leveraging the Independent Causal Mechanism principle, we show that intervention targets can be identified with probability converging exponentially in sample size under faithful Markov boundary recovery and non-degenerate mechanism shifts. Experiments on synthetic benchmarks and five real-world datasets demonstrate that StableRCA is robust to graph misspecification, effective under multiple intervention targets, scalable to large systems, and reliable across diverse application domains. Code is available at: https://anonymous.4open.science/r/StableRCA-E362

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The paper proposes StableRCA, a local mechanism-level root cause analysis framework that estimates Markov boundaries to detect conditional distribution shifts and identify intervention targets without requiring a global causal graph. It claims that, under faithful Markov boundary recovery and non-degenerate mechanism shifts, the probability of correctly identifying intervention targets converges exponentially in sample size, and reports empirical robustness on synthetic benchmarks and five real-world datasets across manufacturing, cloud, and healthcare domains.

Significance. If the stated convergence result holds under the explicitly listed conditions, the work offers a practical graph-agnostic alternative to existing RCA methods that either require accurate global graphs or rely on marginal anomaly detection. The explicit statement of prerequisites (faithful boundary recovery and non-degenerate shifts) and the release of code are positive features that support reproducibility and allow direct testing of the assumptions.

minor comments (3)
  1. The abstract states the exponential convergence result but does not reference the specific theorem or section containing the derivation; adding an explicit pointer (e.g., Theorem 3.2) would improve traceability.
  2. The link to code is given as an anonymous repository; the final version should replace it with a permanent, citable URL or GitHub repository.
  3. Table or figure captions for the real-world datasets should include the number of variables and sample sizes to allow readers to assess scalability claims directly.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive review, accurate summary of our contributions, and recommendation for minor revision. No major comments were raised in the report.

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper's core claim is an exponential convergence result for identifying intervention targets, explicitly conditioned on two external prerequisites (faithful Markov boundary recovery and non-degenerate mechanism shifts) plus the Independent Causal Mechanism principle. These are presented as assumptions rather than derived internally. No equations or steps in the abstract reduce the result to a fitted parameter, self-definition, or self-citation chain; the method treats boundary estimation and shift detection as inputs. This matches the default expectation of a self-contained theoretical statement with no load-bearing circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 3 axioms · 0 invented entities

The central claim rests on the Independent Causal Mechanism principle and the assumptions of faithful Markov boundary recovery plus non-degenerate mechanism shifts; these are invoked directly in the abstract to support the identification guarantee. No free parameters or invented entities are mentioned.

axioms (3)
  • domain assumption Independent Causal Mechanism principle
    Leveraged to establish that intervention targets can be identified via local mechanism shifts.
  • ad hoc to paper Faithful Markov boundary recovery
    Required for the exponential convergence probability result stated in the abstract.
  • ad hoc to paper Non-degenerate mechanism shifts
    Required for the exponential convergence probability result stated in the abstract.

pith-pipeline@v0.9.1-grok · 5728 in / 1393 out tokens · 31761 ms · 2026-06-28T02:26:45.108816+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

83 extracted references · 4 canonical work pages · 1 internal anchor

  1. [1]

    Langley , title =

    P. Langley , title =. 2000 , pages =

    2000

  2. [2]

    T. M. Mitchell. The Need for Biases in Learning Generalizations. 1980

    1980

  3. [3]

    M. J. Kearns , title =

  4. [4]

    Machine Learning: An Artificial Intelligence Approach, Vol. I. 1983

    1983

  5. [5]

    R. O. Duda and P. E. Hart and D. G. Stork. Pattern Classification. 2000

    2000

  6. [6]

    Suppressed for Anonymity , author=

  7. [7]

    Newell and P

    A. Newell and P. S. Rosenbloom. Mechanisms of Skill Acquisition and the Law of Practice. Cognitive Skills and Their Acquisition. 1981

    1981

  8. [8]

    A. L. Samuel. Some Studies in Machine Learning Using the Game of Checkers. IBM Journal of Research and Development. 1959

    1959

  9. [9]

    Causality , publisher=

    Pearl, Judea , year=. Causality , publisher=

  10. [10]

    Proceedings of the IEEE , volume =

    Toward Causal Representation Learning , author =. Proceedings of the IEEE , volume =

  11. [11]

    Proceedings of International Conference on Machine Learning , pages =

    A Theoretical Analysis on Independence-driven Importance Weighting for Covariate-shift Generalization , author =. Proceedings of International Conference on Machine Learning , pages =. 2022 , volume =

    2022

  12. [12]

    2019 , booktitle =

    Shan, Huasong and Chen, Yuan and Liu, Haifeng and Zhang, Yunpeng and Xiao, Xiao and He, Xiaofeng and Li, Min and Ding, Wei , title =. 2019 , booktitle =

    2019

  13. [13]

    2022 , booktitle =

    Ikram, Azam and Chakraborty, Sarthak and Mitra, Subrata and Saini, Shiv Kumar and Bagchi, Saurabh and Kocaoglu, Murat , title =. 2022 , booktitle =

    2022

  14. [14]

    CauseInfer: Automatic and distributed performance diagnosis with hierarchical causality graph in large distributed systems , year=

    Chen, Pengfei and Qi, Yong and Zheng, Pengfei and Hou, Di , booktitle=. CauseInfer: Automatic and distributed performance diagnosis with hierarchical causality graph in large distributed systems , year=

  15. [15]

    2022 , booktitle =

    Li, Mingjie and Li, Zeyan and Yin, Kanglin and Nie, Xiaohui and Zhang, Wenchi and Sui, Kaixin and Pei, Dan , title =. 2022 , booktitle =

    2022

  16. [16]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , pages =

    Li, Jinzhou and Chu, Benjamin B and Scheller, Ines F and Gagneur, Julien and Maathuis, Marloes H , title =. Journal of the Royal Statistical Society Series B: Statistical Methodology , pages =. 2025 , month =

    2025

  17. [17]

    The Annual Conference on Neural Information Processing Systems , year=

    Root Cause Analysis of Outliers with Missing Structural Knowledge , author=. The Annual Conference on Neural Information Processing Systems , year=

  18. [18]

    Proceedings of the AAAI Conference on Artificial Intelligence , year =

    Stable Learning via Sample Reweighting , author =. Proceedings of the AAAI Conference on Artificial Intelligence , year =

  19. [19]

    1988 , publisher=

    Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference , author=. 1988 , publisher=

    1988

  20. [20]

    Lytkin and Jan Lemeire and Constantin F

    Alexander Statnikov and Nikita I. Lytkin and Jan Lemeire and Constantin F. Aliferis , title =. Journal of Machine Learning Research , year =

  21. [21]

    1999 , publisher=

    Practical Nonparametric Statistics , author=. 1999 , publisher=

    1999

  22. [22]

    The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science , volume =

    Karl Pearson , title =. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science , volume =

  23. [23]

    Density Ratio Estimation in Machine Learning , publisher=

    Sugiyama, Masashi and Suzuki, Taiji and Kanamori, Takafumi , year=. Density Ratio Estimation in Machine Learning , publisher=

  24. [24]

    Dvoretzky and J

    A. Dvoretzky and J. Kiefer and J. Wolfowitz , title =. The Annals of Mathematical Statistics , number =

  25. [25]

    Massart , journal =

    P. Massart , journal =. The Tight Constant in the Dvoretzky-Kiefer-Wolfowitz Inequality , volume =

  26. [26]

    2003 , booktitle=

    Inequalities for the L1 Deviation of the Empirical Distribution , author=. 2003 , booktitle=

    2003

  27. [27]

    2024 , eprint=

    The PetShop Dataset -- Finding Causes of Performance Issues across Microservices , author=. 2024 , eprint=

    2024

  28. [28]

    and Peters, Jonas and B

    Gamella, Juan L. and Peters, Jonas and B. Causal chambers as a real-world physical testbed for. Nature Machine Intelligence , year=

  29. [29]

    Automatic root cause analysis in manufacturing: an overview & conceptualization , year =

    e Oliveira, Eduardo and Migu\'. Automatic root cause analysis in manufacturing: an overview & conceptualization , year =. Journal of Intelligent Manufacturing , pages =

  30. [30]

    Survey on Models and Techniques for Root-Cause Analysis

    Marc Solé and Victor Muntés-Mulero and Annie Ibrahim Rana and Giovani Estrada , title =. arXiv:1701.08546 , year =. 1701.08546 , archivePrefix=

    work page internal anchor Pith review Pith/arXiv arXiv

  31. [31]

    and Dimitriou, Nikolaos and Tzovaras, Dimitrios and Margetis, George , title =

    Papageorgiou, Konstantinos and Theodosiou, Theodoros and Rapti, Angeliki and Papageorgiou, Elpiniki I. and Dimitriou, Nikolaos and Tzovaras, Dimitrios and Margetis, George , title =. Frontiers in Manufacturing Technology , volume =

  32. [32]

    ACM Computing Surveys , numpages =

    Soldani, Jacopo and Brogi, Antonio , title =. ACM Computing Surveys , numpages =. 2022 , volume =

    2022

  33. [33]

    and Lipshutz, Alison K

    Wu, Albert W. and Lipshutz, Alison K. M. and Pronovost, Peter J. , title =. Journal of the American Medical Association , volume =

  34. [34]

    and Hettinger, Zane and Shah, Mihir and Wears, Robert L

    Kellogg, Kelly M. and Hettinger, Zane and Shah, Mihir and Wears, Robert L. and Sellers, Charles R. and Squires, Michelle and Fairbanks, Robert J. , title =. BMJ Quality & Safety , volume =. 2017 , month =

    2017

  35. [35]

    Social Science Computer Review , volume =

    Peter Spirtes and Clark Glymour , title =. Social Science Computer Review , volume =

  36. [36]

    Spirtes, Peter and Glymour, Clark and Scheines, Richard , title =

  37. [37]

    2021 , booktitle =

    Varici, Burak and Shanmugam, Karthikeyan and Sattigeri, Prasanna and Tajer, Ali , title =. 2021 , booktitle =

    2021

  38. [38]

    BACKSHIFT: Learning causal cyclic graphs from unknown shift interventions , year =

    Rothenh\". BACKSHIFT: Learning causal cyclic graphs from unknown shift interventions , year =. Proceedings of the 29th International Conference on Neural Information Processing Systems - Volume 1 , pages =

  39. [39]

    Proceedings of Conference on Uncertainty in Artificial Intelligence , year=

    Intervention Target Estimation in the Presence of Latent Variables , author=. Proceedings of Conference on Uncertainty in Artificial Intelligence , year=

  40. [40]

    2023 , booktitle =

    Wang, Dongjie and Chen, Zhengzhang and Fu, Yanjie and Liu, Yanchi and Chen, Haifeng , title =. 2023 , booktitle =

    2023

  41. [41]

    2024 , booktitle =

    Zheng, Lecheng and Chen, Zhengzhang and He, Jingrui and Chen, Haifeng , title =. 2024 , booktitle =

    2024

  42. [42]

    2024 , booktitle =

    Lin, Cheng-Ming and Chang, Ching and Wang, Wei-Yao and Wang, Kuang-Da and Peng, Wen-Chih , title =. 2024 , booktitle =

    2024

  43. [43]

    Journal of Systems and Software , numpages =

    Xin, Ruyue and Chen, Peng and Zhao, Zhiming , title =. Journal of Systems and Software , numpages =. 2023 , volume =

    2023

  44. [44]

    Microscope: Pinpoint Performance Issues with Causal Graphs in Micro-service Environments

    Lin, Jinjin and Chen, Pengfei and Zheng, Zibin. Microscope: Pinpoint Performance Issues with Causal Graphs in Micro-service Environments. Service-Oriented Computing. 2018

    2018

  45. [45]

    2020 , booktitle =

    Ma, Meng and Xu, Jingmin and Wang, Yuan and Chen, Pengfei and Zhang, Zonghua and Wang, Ping , title =. 2020 , booktitle =

    2020

  46. [46]

    2024 , booktitle =

    Pham, Luan and Ha, Huong and Zhang, Hongyu , title =. 2024 , booktitle =

    2024

  47. [47]

    2024 , booktitle =

    Nguyen, Phuoc and Tran, Truyen and Gupta, Sunil and Nguyen, Thin and Venkatesh, Svetha , title =. 2024 , booktitle =

    2024

  48. [48]

    The International Conference on Learning Representations , year=

    Robust Root Cause Diagnosis using In-Distribution Interventions , author=. The International Conference on Learning Representations , year=

  49. [49]

    2023 , booktitle =

    Han, Xiao and Zhang, Lu and Wu, Yongkai and Yuan, Shuhan , title =. 2023 , booktitle =

    2023

  50. [50]

    and Margaritis, D

    Yaramakala, S. and Margaritis, D. , booktitle=. Speculative Markov blanket discovery for optimal feature selection , year=

  51. [51]

    Accurate Markov Boundary Discovery for Causal Feature Selection , year=

    Wu, Xingyu and Jiang, Bingbing and Yu, Kui and Miao, chunyan and Chen, Huanhuan , journal=. Accurate Markov Boundary Discovery for Causal Feature Selection , year=

  52. [52]

    Bayesian Network Induction via Local Neighborhoods , volume =

    Margaritis, Dimitris and Thrun, Sebastian , booktitle =. Bayesian Network Induction via Local Neighborhoods , volume =

  53. [53]

    and Statnikov, Alexander , title =

    Tsamardinos, Ioannis and Aliferis, Constantin F. and Statnikov, Alexander , title =. Proceedings of International Florida Artificial Intelligence Research Society Conference , year =

  54. [54]

    Learning Gaussian Graphical Models of Gene Networks with False Discovery Rate Control , booktitle =

    Pe. Learning Gaussian Graphical Models of Gene Networks with False Discovery Rate Control , booktitle =. 2008 , pages =

    2008

  55. [55]

    Proceedings of Conference on Uncertainty in Artificial Intelligence , pages =

    Learning LWF Chain Graphs: A Markov Blanket Discovery Approach , author =. Proceedings of Conference on Uncertainty in Artificial Intelligence , pages =. 2020 , volume =

    2020

  56. [56]

    1996 , booktitle =

    Koller, Daphne and Sahami, Mehran , title =. 1996 , booktitle =

    1996

  57. [57]

    Conditional Likelihood Maximisation: A Unifying Framework for Information Theoretic Feature Selection , journal =

    Gavin Brown and Adam Pocock and Ming-Jie Zhao and Mikel Luj. Conditional Likelihood Maximisation: A Unifying Framework for Information Theoretic Feature Selection , journal =. 2012 , volume =

    2012

  58. [58]

    Stable Prediction with Model Misspecification and Agnostic Distribution Shift , volume=

    Kuang, Kun and Xiong, Ruoxuan and Cui, Peng and Athey, Susan and Li, Bo , year=. Stable Prediction with Model Misspecification and Agnostic Distribution Shift , volume=. Proceedings of the AAAI Conference on Artificial Intelligence , pages=

  59. [59]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , month =

    Zhang, Xingxuan and Cui, Peng and Xu, Renzhe and Zhou, Linjun and He, Yue and Shen, Zheyan , title =. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , month =. 2021 , pages =

    2021

  60. [60]

    , title =

    Quinonero-Candela, Joaquin and Sugiyama, Masashi and Schwaighofer, Anton and Lawrence, Neil D. , title =. 2009 , pages =

    2009

  61. [61]

    2012 , month =

    Sugiyama, Masashi and Kawanabe, Motoaki , title =. 2012 , month =

    2012

  62. [62]

    and Bagchi, Saurabh and Inouye, David I

    Kulinski, Sean M. and Bagchi, Saurabh and Inouye, David I. , title =. 2020 , booktitle =

    2020

  63. [63]

    Kernel measures of conditional dependence , year =

    Fukumizu, Kenji and Gretton, Arthur and Sun, Xiaohai and Sch\". Kernel measures of conditional dependence , year =. Proceedings of International Conference on Neural Information Processing Systems , pages =

  64. [64]

    Shimodaira, Hidetoshi , journal =

  65. [65]

    2018 , booktitle =

    Prokhorenkova, Liudmila and Gusev, Gleb and Vorobev, Aleksandr and Dorogush, Anna Veronika and Gulin, Andrey , title =. 2018 , booktitle =

    2018

  66. [66]

    Covariate Shift Adaptation by Importance Weighted Cross Validation , year =

    Sugiyama, Masashi and Krauledat, Matthias and M\". Covariate Shift Adaptation by Importance Weighted Cross Validation , year =. Journal of Machine Learning Research , month = dec, pages =

  67. [67]

    ProRCA: A Causal Python Package for Actionable Root Cause Analysis in Real-world Business Scenarios , journal =

    Ahmed Dawoud and Shravan Talupula , year=. ProRCA: A Causal Python Package for Actionable Root Cause Analysis in Real-world Business Scenarios , journal =. 2503.01475 , archivePrefix=

    work page arXiv

  68. [68]

    Proceedings of Conference on Uncertainty in Artificial Intelligence , year=

    Extremely Greedy Equivalence Search , author=. Proceedings of Conference on Uncertainty in Artificial Intelligence , year=

  69. [69]

    2022 , howpublished =

    Sock-shop: A Microservice Demo Application , author =. 2022 , howpublished =

    2022

  70. [70]

    Companion Proceedings of the ACM Web Conference 2025 , year =

    Pham, Luan and Zhang, Hongyu and Ha, Huong and Salim, Flora and Zhang, Xiuzhen , title =. Companion Proceedings of the ACM Web Conference 2025 , year =

    2025

  71. [71]

    and Mendelson, Shahar , title =

    Bartlett, Peter L. and Mendelson, Shahar , title =. Journal of Machine Learning Research , month = mar, pages =. 2003 , issue_date =

    2003

  72. [72]

    arXiv:2502.12707 , year=

    CausalMan: A physics-based simulator for large-scale causality , author=. arXiv:2502.12707 , year=. 2502.12707 , archivePrefix=

    work page arXiv

  73. [73]

    Publicationes Mathematicae Debrecen , pages = 290, title =

    Erd\". Publicationes Mathematicae Debrecen , pages = 290, title =

  74. [74]

    International conference on machine learning , pages=

    Causal structure-based root cause analysis of outliers , author=. International conference on machine learning , pages=. 2022 , organization=

    2022

  75. [75]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , pages=

    Root cause discovery via permutations and Cholesky decomposition , author=. Journal of the Royal Statistical Society Series B: Statistical Methodology , pages=. 2025 , publisher=

    2025

  76. [76]

    The World Wide Web Conference , pages=

    ?-diagnosis: Unsupervised and real-time diagnosis of small-window long-tail latency in large-scale microservice platforms , author=. The World Wide Web Conference , pages=

  77. [77]

    IEEE/ACM International Conference on Software Engineering: Software Engineering in Practice , pages=

    Microhecl: High-efficient root cause localization in large-scale microservice systems , author=. IEEE/ACM International Conference on Software Engineering: Software Engineering in Practice , pages=

  78. [78]

    Cumulated gain-based evaluation of IR techniques , year =

    J\". Cumulated gain-based evaluation of IR techniques , year =. ACM Transactions on Information Systems , month = oct, pages =

  79. [79]

    2018 18th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID) , pages =

    Ping Wang and Jingmin Xu and Meng Ma and Weilan Lin and Disheng Pan and Yuan Wang and Pengfei Chen , title =. 2018 18th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID) , pages =. 2018 , publisher =

    2018

  80. [80]

    Proceedings of the Forty-First Conference on Uncertainty in Artificial Intelligence , volume =

    Root Cause Analysis of Failures from Partial Causal Structures , author =. Proceedings of the Forty-First Conference on Uncertainty in Artificial Intelligence , volume =

Showing first 80 references.