Pith. sign in

REVIEW 4 major objections 6 minor 48 references

Interpretable Event Diagnosis in Water Distribution Networks

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Counterfactual fingerprints can separate leaks from sensor faults in water distribution networks.

desk verdict The paper's core idea is fine, but the CDF formalism doesn't match the reported fingerprints—following Eq. (11) can't produce the sensor-fault results, so the empirical claims aren't reproducible as written. read the letter →

arxiv 2505.07299 v1 pith:FZWTFVIR submitted 2025-05-12 cs.AI cs.LGcs.SYeess.SY

André Artelt , Stelios G. Vrachimis , Demetrios G. Eliades , Ulrike Kuhl , Barbara Hammer , Marios M. Polycarpou This is my paper · ORCID
classification cs.AIcs.LGcs.SYeess.SY
keywords waterdistributionnetworkseventdiagnosiscounterfactualexplanationsinterpretablemachinelearningleakagedetectionsensorfaultisolationfingerprints
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to establish that a 'counterfactual event detection fingerprint'—the smallest set of changes to current sensor readings that would make a detected alarm disappear—is a useful and interpretable representation for diagnosing events in water distribution networks. The authors claim that leaks and sensor faults leave different fingerprint patterns: a sensor fault lights up mostly the faulty sensor, while a leak spreads changes across several sensors. On that basis, a decision tree trained on fingerprints distinguishes leakages from sensor faults with F1 scores of 0.94 for both classes on the Hanoi network and 0.97 for both on the L-Town network, outperforming a residual-based baseline especially for small leaks. The motivation is operator trust: instead of a black-box alarm, the operator sees what would have to change for the alarm to go away, which can be compared with their own experience. If correct, the framework turns event diagnosis into an inspectable, feedback-friendly loop rather than an opaque classification output.

What carries the argument

The central object is the counterfactual event detection fingerprint (CDF): a vector $\delta_{\mathrm{CDF}}$ that, added to the observed sensor window, would flip the event-detection function from 'alarm' to 'no alarm', chosen to be closest to the observed data in the $\ell^1$ norm and constrained to lie among recorded normal-operation measurements. It is computed as an ensemble-consistent explanation, meaning it must silence all virtual sensors in the detection ensemble at once. The companion counterfactual event isolation fingerprint (CIF) is the analogous minimal change that would flip the event-type classifier's output. The fingerprint carries the argument because its pattern of nonzero entries—spread out for leaks, concentrated on one sensor for faults—is what separates the two event classes and gives the operator a visual handle on the algorithm's decision.

What would settle it

Show experienced water-network operators alarm logs from real events both with and without counterfactual fingerprint plots, and compare their event-type judgments: if the plots do not measurably improve accuracy or confidence, the claimed interpretability benefit is unsupported. For the classification claim itself, applying the same CDF decision-tree pipeline to a third, independent water network would reveal whether the leak-versus-sensor-fault separation is a general property or an artifact of the two benchmark networks.

Watch

Extended reading notes

Core claim

The central claim is that leakages and sensor faults produce robustly different counterfactual event detection fingerprints, and that these fingerprints are sufficient input for an interpretable classifier to identify the event type. Formally, given a residual-based detector that raises an alarm when a virtual-sensor forecast differs from observations, the closest CDF is the minimal change to the observed sensor window, restricted to plausible normal-operation measurements, that would turn the alarm off. The paper observes that leak CDFs tend to involve coordinated changes at many sensors, whereas sensor-fault CDFs concentrate on the faulty sensor; this contrast is what makes the downstream classification work. In the two case studies, the CDF-based decision tree reaches F1 0.94 for leaks and 0.94 for sensor faults in Hanoi and F1 0.97 for both classes in L-Town, while the direct-residual decision-tree baseline scores 0.70 on Hanoi leaks and 0.29 on Hanoi sensor faults. The authors also define counterfactual event isolation fingerprints (CIFs), which show what would have to change for the classifier to switch its diagnosis, extending the same explanation idea to the isolation stage.

Load-bearing premise

The load-bearing assumption is that operators will actually find counterfactual fingerprint plots intuitive and actionable; the paper does not test whether human understanding, trust, or decision quality improves when these explanations are shown.

Editorial extensions

If this is right

  • If the fingerprints really separate leaks from sensor faults, the same CDF computation can be reused to explain detection for any residual-based event detector, not only the linear-regression virtual sensors tested here.
  • Small leaks, which are the hardest cases for residual baselines, become classifiable from fingerprints: the L-Town small-leak F1 rises from 0.48 for the decision-tree baseline to 0.93 for the CDF-based decision tree.
  • Because the isolation classifier is a decision tree on fingerprints, the full pipeline remains inspectable: an operator can see both why an alarm fired and why a particular class label was assigned.
  • CIFs provide a concrete way to ask 'what would change this diagnosis?', which can support operator review, hypothetical reasoning, and feedback labeling in the monitoring loop.
  • The authors' stated boundary is that the method only handles event types present in training data; unknown event types are left for future work.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper's motivating claim about operator trust is untested: no experiment measures whether operators actually understand fingerprint plots, trust them, or make better decisions with them; that would require a user study.
  • The discriminative power of the fingerprint likely depends on the residual detector's structure and on sensor placement; a network with very few or poorly placed sensors might not show the same leak-versus-fault contrast.
  • The same sparsity pattern that separates classes could be pushed further, for example to localize the leaking node or the faulty sensor, since a leak's fingerprint appears to emphasize sensors near the event.
  • Restricting fingerprints to recorded normal-operation data makes the explanation depend on the training period's demand patterns; unusual but legitimate hydraulic states could degrade both the explanation's plausibility and the downstream classification.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes a framework for interpretable event diagnosis in water distribution networks. The authors introduce counterfactual event detection fingerprints (CDFs) and counterfactual event isolation fingerprints (CIFs), which are computed by finding the closest alternative sensor measurement pattern that would change an alarm or a classification decision. The CDFs are then used as features to train an interpretable classifier (decision tree or k-NN) to distinguish leakages from sensor faults. The methodology is evaluated on the Hanoi benchmark and on a modified L-Town network instance, reporting F1 scores around 0.94-0.97 for the proposed method, outperforming a residual-based baseline for small leakages. The paper also discusses limitations and future work, including the need for a user-centric evaluation.

Significance. If the results are reproducible and the definitions are tightened, the paper would make a valuable contribution to XAI for critical infrastructure: it is, to the authors' knowledge, the first systematic proposal of counterfactual fingerprints for WDN event diagnosis, and the experimental design (training on a simulated model and testing on a different network instance) is a genuine generalization test. The proposed representation is simple, visual, and potentially actionable for operators. However, the current manuscript contains a load-bearing gap in the definition of the CDF optimization (Eq. 11), which appears infeasible for the sensor-fault cases that are central to the evaluation, and the detection metrics in Tables 1 and 4 are internally inconsistent. The interpretability claim is also not yet supported by user evidence, though the authors explicitly acknowledge this.

major comments (4)
  1. [Section 6, Eq. (11)] The closest-CDF optimization as stated is infeasible for typical sensor faults. Since f_i is a linear regression that ignores sensor i (Section 5, Eq. (7)), the constraint |f_i(xcf) - y(k)_i| <= theta_i cannot be influenced by xcf_i. When y(k)_i is a power failure value (0 m) or a large offset while all other sensors are normal, any xcf in the normal training set D yields f_i(xcf) near the normal pressure range, so the constraint cannot be satisfied. The paper does not specify how infeasible cases are handled, and Fig. 4(a) shows the largest fingerprint change on the faulty sensor, which is exactly the component that Eq. (11) cannot use to satisfy the constraint for that sensor. Therefore, either the implemented CDF differs from Definition 3/Eq. (11), or the reported sensor-fault fingerprints and the corresponding classification results (Tables 2 and 5) are not produced by the stated algorithm. The authors should clarify the algorithm actually used and provide a feasibility analysis.
  2. [Section 6, Definition 2 vs. Eqs. (10)-(12)] The formal definition of a CDF as a perturbation of the observed time window is inconsistent with the implementation, which solves for a new previous measurement vector xcf = y(k-1) + delta* while holding y(k) fixed as the target. Consequently, the CDF plotted in Fig. 4 represents changes to the previous time step, not to the measurements at the moment of detection as the text and caption suggest. This should be stated explicitly, and Definition 2 and Eq. (9) should be rewritten to match the actual computational problem, including the T=1 specialization.
  3. [Tables 1 and 4] The reported detection rates violate basic identities: TP+FN must equal 1 and FP+TN must equal 1 for rates computed on the same event/non-event bases. In Table 1, Leakage shows 0.98+0.08=1.06 and Small shows 0.96+0.22=1.18; similar inconsistencies appear in Table 4. The authors should report the exact definitions of TP, FP, FN, and TN (e.g., per time step vs. per event) and ensure the numbers are consistent.
  4. [Sections 1, 7, and 11] The interpretability claim is not empirically validated. The paper acknowledges in Section 11 that 'A user-centric evaluation will be instrumental in demonstrating the methodology's real-world applicability,' but the abstract and title present interpretability as an achieved property. At minimum, the claims should be scoped to 'representation interpretability,' or a study with operators should be added. Without this, the central value proposition relative to non-interpretable baselines is not demonstrated.
minor comments (6)
  1. [Eq. (7)] The summation index j is not defined consistently with the argument y(k-i); there is a typo that should be corrected.
  2. [Eq. (8) vs. Eq. (11)] Eq. (8) uses a single threshold theta, while Eq. (11) uses per-sensor thresholds theta_i; the relationship between these thresholds should be clarified.
  3. [Section 7] The baseline is called 'Non-interpretable baseline,' but a decision tree is itself interpretable; the distinction is in the input representation (residuals vs. CDFs), not the model class, so the label is misleading.
  4. [Figures 4 and 6] The color scales differ between the subfigures within each figure; a shared scale would make the comparison more direct.
  5. [Section 10.1] The number of test leakage scenarios (35) is much smaller than the training set; the authors should report confidence intervals or acknowledge this limitation explicitly.
  6. [Tables 1 and 4] The text says 'mean an variance' but reports values with +/- that appear to be standard deviations; the notation should be made precise.

Circularity Check

0 steps flagged · score 0.0 of 10

The derivation is self-contained: the isolation classifier is trained on counterfactual fingerprints from simulated scenarios and tested on a separate network instance, so no prediction reduces to a fitted input.

full rationale

The paper's derivation chain is not circular. The CDF is defined by a concrete optimization (Definition 3 and Eq. (11)) over the normal-operation training set, and the event-isolation classifier is trained on CDFs from simulated scenarios and evaluated on a different L-Town instance with realistic demands and changed hydraulics (Section 10.1). The claim that leak and sensor-fault CDFs differ is an empirical observation from the case studies (Figs. 4 and 6), not an identity forced by the definitions; the modeling assumption that a sensor fault affects one sensor while a leak affects several does not by itself prove the reported F1 scores. The self-citation [47] only supplies terminology ('ensemble consistent explanation') for an optimization that is fully specified in Eqs. (10)-(12), so it is not load-bearing. The acknowledged absence of a user study (Section 11) is a limitation of the interpretability claim, not a circular dependency. The possible infeasibility of Eq. (11) for some sensor-fault conditions raised in review is a correctness/reproducibility concern rather than a circularity, and it does not alter this score.

Assumptions & free parameters 4 free parameters · 5 assumptions · 2 invented entities

The paper is an applied XAI method, not a first-principles derivation. The central results rest on standard hydraulic modeling assumptions, a linear-regression virtual sensor approximation, and the untested premise that operators will find the fingerprints useful. Free parameters include detection thresholds, window length, regularization, and classifier hyperparameters, none of which are fully specified.

free parameters (4)
  • detection threshold theta = not reported
    Estimated on a fault-free training period as slightly larger than the maximum observed residual (Section 5); a free parameter that controls detection sensitivity and therefore all CDFs.
  • time window length T = 1
    Set to 1 'as this will be sufficient in practice' (Section 6); a hand-chosen model choice that limits fingerprints to single time steps.
  • least-squares regularization strength = not reported
    The virtual sensors use regularized least-squares (Section 5), but the regularization coefficient is unspecified.
  • decision tree hyperparameters or k in k-NN = not reported
    Used for event isolation classifiers (Sections 7, 9, 10); no tuning procedure or final values are given.
assumptions (5)
  • domain assumption Standard hydraulic conservation equations (mass and energy) describe WDN behavior
    Section 3, Eqs. (2)-(3), cited from [42]; underlies the simulation of events and the digital model used to train the classifier.
  • domain assumption The network operates in pressure-sufficient conditions and reservoir/tank heads are known
    Stated explicitly in Section 3 as modeling assumptions of Eq. (3).
  • domain assumption Leakage is pressure-dependent according to Eq. (5) with emitter coefficient c_j and exponent alpha
    Section 3, from [43]; used to simulate leakages.
  • domain assumption Linear regression virtual sensors (Eq. 7) provide adequate forecasts of sensor values from other sensors
    Section 5; the entire residual-based detection and CDF computation rest on this approximation; no comparison with nonlinear models is made in this work.
  • ad hoc to paper CDF/CIF explanations are plausible and intuitive for operators
    The claimed value proposition assumes operators will understand and trust the fingerprints; the paper provides no user study and acknowledges this in Section 11.
invented entities (2)
  • Counterfactual Event Detection Fingerprint (CDF)
    purpose: Vector of sensor-measurement changes that would revert an alarm to no-alarm; used as an interpretable explanation and as a feature for event classification
    Defined in Section 6; its usefulness is only evaluated through downstream classification accuracy, not through independent behavioral or cognitive evidence.
  • Counterfactual Event Isolation Fingerprint (CIF)
    purpose: Vector of changes to a CDF that would flip the event-type classification; intended to explain isolation decisions
    Defined in Section 8; no independent evaluation is provided (no illustrative or quantitative results for CIFs are shown).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Interpretable Event Diagnosis in Water Distribution Networks." pith.science (2026). https://pith.science/paper/FZWTFVIR

@misc{pith2026250507299,
  author       = {Pith},
  title        = {Pith review of: Interpretable Event Diagnosis in Water Distribution Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FZWTFVIR}},
  note         = {Machine review of arXiv:2505.07299}
}
read the original abstract

The increasing penetration of information and communication technologies in the design, monitoring, and control of water systems enables the use of algorithms for detecting and identifying unanticipated events (such as leakages or water contamination) using sensor measurements. However, data-driven methodologies do not always give accurate results and are often not trusted by operators, who may prefer to use their engineering judgment and experience to deal with such events. In this work, we propose a framework for interpretable event diagnosis -- an approach that assists the operators in associating the results of algorithmic event diagnosis methodologies with their own intuition and experience. This is achieved by providing contrasting (i.e., counterfactual) explanations of the results provided by fault diagnosis algorithms; their aim is to improve the understanding of the algorithm's inner workings by the operators, thus enabling them to take a more informed decision by combining the results with their personal experiences. Specifically, we propose counterfactual event fingerprints, a representation of the difference between the current event diagnosis and the closest alternative explanation, which can be presented in a graphical way. The proposed methodology is applied and evaluated on a realistic use case using the L-Town benchmark.

Figures

Figures reproduced from arXiv: 2505.07299 by the authors.

Figure 1
Figure 1. Illustration of the proposed Interpretable Event Diagnosis methodology: The blue block indicates [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. L-Town-Network (∆t = 5min): Pressure forecasts in normal vs. leaky times – a leakage is present from time 170 onwards (starting point indicated by the vertical dashed red line). 6. Counterfactual Event Detection Fingerprints (CDF) In the following, we formalize the concept of counterfactual event detection fingerprints which are used to explain the event detection and, as a downstream task, to classify (i.e. determi… view at source ↗
Figure 3
Figure 3. The Hanoi network including four (4) pressure sensors and one (1) flow sensor at the inlet. is monitored, i.e., the head at the reservoir node 1 and the flow at link 1 are measured. Ad￾ditionally, 4 pressure sensors are installed at nodes {13, 16, 22, 30} using a sensor placement procedure that maximizes the minimum sensitivity of all sensors to all possible leakages [48]. The Event Detection module (see Section 5) … view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Hanoi Network: Illustration of CDFs (Definition 3) of sensor faults vs. leakages – the scale [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]
Figure 5
Figure 5. Figure 5: The L-Town network (Area A) (Vrachimis et al., 2022) and the locations of [PITH_FULL_IMAGE:figures/full_fig_p017_5.png]
Figure 6
Figure 6. Figure 6: L-Town network: Illustration of CDFs (Definition 3) of sensor faults vs. leakages – the scale [PITH_FULL_IMAGE:figures/full_fig_p018_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 35 canonical work pages

  1. [1]

    Giustolisi, D

    O. Giustolisi, D. Savic, Z. Kapelan, Pressure-Driven Demand and Leakage Simulation for Water Distribution Networks, Journal of Hydraulic Engineering 134 (5) (2008) 626–

    work page 2008

  2. [2]

    D. G. Eliades, M. M. Polycarpou, Leakage fault detection in district metered areas of water distribution systems, Journal of Hydroinformatics 14 (4) (2012) 992–1005. doi:10.2166/hydro.2012.109

    work page doi:10.2166/hydro.2012.109 2012

  3. [3]

    M. À. Cugueró-Escofet, V. Puig, J. Quevedo, Optimal pressure sensor place- ment and assessment for leak location using a relaxed isolation index: Applica- tion to the Barcelona water network, Control Engineering Practice 63 (2017) 1–12. doi:10.1016/j.conengprac.2017.03.003. 22

    work page doi:10.1016/j.conengprac.2017.03.003 2017

  4. [4]

    Reppa, M

    V. Reppa, M. M. Polycarpou, C. G. Panayiotou, Sensor Fault Diagnosis, Foundations and Trends®in Systems and Control 3 (1-2) (2016) 1–248. doi:10.1561/2600000007. URLhttp://www.nowpublishers.com/article/Details/SYS-007

    work page doi:10.1561/2600000007 2016

  5. [5]

    Pecci, P

    F. Pecci, P. Parpas, I. Stoianov, Sequential Convex Optimization for Detecting and Lo- cating Blockages in Water Distribution Networks, Journal of Water Resources Planning and Management 146 (8) (2020) 04020057. doi:10.1061/(ASCE)WR.1943-5452.0001233. URLhttp://ascelibrary.org/doi/10.1061/%28ASCE%29WR.1943-5452.0001233

    arXiv 2020

  6. [6]

    S. G. Vrachimis, S. Timotheou, D. G. Eliades, M. M. Polycarpou, Leakage detection and localization in water distribution systems: A model invalidation approach, Control Engineering Practice 110 (2021) 104755. doi:10.1016/j.conengprac.2021.104755. URLhttps://linkinghub.elsevier.com/retrieve/pii/S0967066121000320

    arXiv 2021

  7. [7]

    Zou, Y.-L

    X.-Y. Zou, Y.-L. Lin, B. Xu, Z.-B. Guo, S.-J. Xia, T.-Y. Zhang, A.-Q. Wang, N.-Y. Gao, A Novel Event Detection Model for Water Distribution Systems Based on Data-Driven Estimation and Support Vector Machine Classification, Water Resources Management (nov 2019). doi:10.1007/s11269-019-02317-5. URLhttp://link.springer.com/10.1007/s11269-019-02317-5

    work page doi:10.1007/s11269-019-02317-5 2019

  8. [8]

    Soldevila, J

    A. Soldevila, J. Blesa, T. N. Jensen, S. Tornil-Sin, R. M. Fernandez-Canti, V. Puig, Leak Localization Method for Water-Distribution Networks Using a Data-Driven Model and Dempster-Shafer Reasoning, IEEE Transactions on Control Systems Technology 29 (3) (2021) 937–948. doi:10.1109/TCST.2020.2982349

    work page arXiv 2021

Show all 48 references
  1. [9]

    L. Peng, J. Zhang, Y. Li, G. Du, A novel percussion-based approach for pipeline leakage detection with improved mobilenetv2, Engineering Applications of Artificial Intelligence 133 (2024) 108537. doi:10.1016/j.engappai.2024.108537

    2024

  2. [10]

    M. R. Islam, S. Azam, B. Shanmugam, D. Mathur, Leak detection and localiza- tion in underground water supply system using thermal imaging and geophone sig- nals through machine learning, Intelligent Systems with Applications 23 (2024) 200404. doi:https://doi.org/10.1016/j.iswa...

    2024

  3. [11]

    X. Zhou, Z. Tang, W. Xu, F. Meng, X. Chu, K. Xin, G. Fu, Deep learning identifies accurate burst locations in water distribution networks, Water Research 166 (dec 2019). doi:10.1016/j.watres.2019.115058

    2019

  4. [13]

    Y. Wu, S. Liu, A review of data-driven approaches for burst detection in water distri- bution systems (oct 2017). doi:10.1080/1573062X.2017.1279191. 23

    2017

  5. [14]

    Kahneman, J

    D. Kahneman, J. L. Knetsch, R. H. Thaler, Anomalies: The Endowment Effect, Loss Aversion, and Status Quo Bias, Journal of Economic Perspectives 5 (1) (1991) 193–206. doi:10.1257/JEP.5.1.193

    1991 doi

  6. [15]

    A. M. Mustafa, R. Agha, L. Ghazalat, T. Sha’ban, Natural disasters detection us- ing explainable deep learning, Intelligent Systems with Applications 23 (2024) 200430. doi:https://doi.org/10.1016/j.iswa.2024.200430. URLhttps://www.sciencedirect.com/science/article/pii/S2667305...

    2024

  7. [16]

    European Commission, White Paper on Artificial Intelligence: a European approach to excellence and trust, AI for Europe, COM/2018/237 final (19-02-2020)

    2018

  8. [17]

    Larsson, F

    S. Larsson, F. Heintz, Transparency in artificial intelligence, Internet Policy Review 9 (2) (2020). doi:DOI:10.14763/2020.2.1469

    2020 doi

  9. [18]

    Council of European Union, Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC, Offic...

    2016

  10. [19]

    EuropeanCommissionandDirectorate-GeneralforCommunicationsNetworksandCon- tent and Technology, Proposal for a Regulation laying down harmonised rules on Ar- tificial Intelligence (Artificial Intelligence Act) and amending certain Union legislative acts, Policy and Legistlation ...

    2021

  11. [20]

    Miller, Explanation in artificial intelligence: Insights from the social sciences, Arti- ficial intelligence 267 (2019) 1–38

    T. Miller, Explanation in artificial intelligence: Insights from the social sciences, Arti- ficial intelligence 267 (2019) 1–38

    2019

  12. [21]

    Guidotti, A

    R. Guidotti, A. Monreale, S. Ruggieri, F. Turini, F. Giannotti, D. Pedreschi, A survey of methods for explaining black box models, ACM Comput. Surv. 51 (5) (Aug. 2018)

    2018

  13. [22]

    Molnar, Interpretable Machine Learning, Leanpub, 2019

    C. Molnar, Interpretable Machine Learning, Leanpub, 2019

    2019

  14. [23]

    I know it when I see it

    F. Offert, “I know it when I see it”. Visualization and Intuitive Interpretability, arXiv (2017). arXiv:1711.08042

    2017 arXiv

  15. [24]

    K. J. Rohlfing, P. Cimiano, I. Scharlau, T. Matzner, H. M. Buhl, H. Buschmeier, E. Es- posito, A. Grimminger, B. Hammer, R. Häb-Umbach, I. Horwath, E. Hüllermeier, F. Kern, S. Kopp, K. Thommes, A. N. Ngomo, C. Schulte, H. Wachsmuth, P. Wag- ner, B. Wrede, Explanation as a soci...

    2021

  16. [25]

    why should i trust you?

    M. T. Ribeiro, S. Singh, C. Guestrin, "why should i trust you?": Explaining the pre- dictions of any classifier, in: Proceedings of the 22Nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD ’16, ACM, New York, NY, USA, 2016, pp. 1135–1144. doi:10...

    2016

  17. [26]

    S. Bach, A. Binder, G. Montavon, F. Klauschen, K.-R. Müller, W. Samek, On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation, PloS one 10 (7) (2015) e0130140

    2015

  18. [27]

    Fisher, C

    A. Fisher, C. Rudin, F. Dominici, All Models are Wrong but many are Useful: Variable Importance for Black-Box, Proprietary, or Misspecified Prediction Models, using Model Class Reliance, arXiv e-prints (2018) arXiv:1801.01489arXiv:1801.01489

    2018 arXiv

  19. [28]

    B. Kim, O. Koyejo, R. Khanna, Examples are not enough, learn to criticize! criticism for interpretability, in: Advances in Neural Information Processing Systems 29, 2016

    2016

  20. [29]

    Wachter, B

    S. Wachter, B. Mittelstadt, C. Russell, Counterfactual explanations without opening the black box: Automated decisions and the gdpr, Harv. JL & Tech. 31 (2017) 841

    2017

  21. [30]

    Verma, J

    S. Verma, J. Dickerson, K. Hines, Counterfactual explanations for machine learning: A review (2020). arXiv:2010.10596

    2020 arXiv

  22. [31]

    R. M. J. Byrne, Counterfactuals in explainable artificial intelligence (xai): Evidence from human reasoning, in: IJCAI-19, 2019

    2019

  23. [32]

    Guidotti, Counterfactual explanations and how to find them: literature review and benchmarking, Data Mining and Knowledge Discovery (2022) 1–55

    R. Guidotti, Counterfactual explanations and how to find them: literature review and benchmarking, Data Mining and Knowledge Discovery (2022) 1–55

    2022

  24. [33]

    Poyiadzi, K

    R. Poyiadzi, K. Sokol, R. Santos-Rodriguez, T. De Bie, P. Flach, Face: feasible and actionable counterfactual explanations, in: Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, 2020, pp. 344–350

    2020

  25. [34]

    Chakraborty, M

    A. Chakraborty, M. Alam, V. Dey, A. Chattopadhyay, D. Mukhopadhyay, A survey on adversarial attacks and defences, CAAI Transactions on Intelligence Technology 6 (1) (2021) 25–45

    2021

  26. [35]

    Artelt, B

    A. Artelt, B. Hammer, Convex optimization for actionable \& plausible counterfactual explanations, CoRR abs/2105.07630 (2021). arXiv:2105.07630. URLhttps://arxiv.org/abs/2105.07630

    2021 arXiv

  27. [36]

    Van Looveren, J

    A. Van Looveren, J. Klaise, Interpretable counterfactual explanations guided by proto- types (2021) 650–665

    2021

  28. [37]

    Artelt, B

    A. Artelt, B. Hammer, Convex density constraints for computing plausible coun- terfactual explanations, 29th International Conference on Artificial Neural Networks (ICANN), 2020. doi:10.1007/978-3-030-61609-0_28

    2020 doi

  29. [38]

    R. K. Mothilal, A. Sharma, C. Tan, Explaining machine learning classifiers through diverse counterfactual explanations, in: Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, 2020, pp. 607–617

    2020

  30. [39]

    Russell, Efficient search for diverse coherent explanations, in: Proceedings of the Conference on Fairness, Accountability, and Transparency, 2019, pp

    C. Russell, Efficient search for diverse coherent explanations, in: Proceedings of the Conference on Fairness, Accountability, and Transparency, 2019, pp. 20–28. 25

    2019

  31. [40]

    Rodriguez, M

    P. Rodriguez, M. Caccia, A. Lacoste, L. Zamparo, I. Laradji, L. Charlin, D. Vazquez, Beyond trivial counterfactual explanations with diverse valuable explanations, in: Pro- ceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 1056–1065

    2021

  32. [41]

    why here and not there?

    A. Artelt., A. Schulz., B. Hammer., "why here and not there?": Diverse contrasting explanations of dimensionality reduction, in: Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - ICPRAM„ INSTICC, SciTePress, 2023, pp. 27–38. doi...

    2023 doi

  33. [42]

    S. G. Vrachimis, S. Timotheou, D. G. Eliades, M. M. Polycarpou, Iterative hydraulic interval state estimation for water distribution networks, Journal of Water Resources Planning and Management 145 (1) (2018) 04018087. doi:10.1061/(ASCE)WR.1943- 5452.0001011

    2018

  34. [43]

    A. M. Kabaasha, J. E. van Zyl, G. K. Mahinthakumar, Correcting power leakage equa- tion for improved leakage modeling and detection, Journal of Water Resources Planning and Management 146 (3) (2020) 06020001. doi:10.1061/(ASCE)WR.1943-5452.0001172

    2020

  35. [44]

    Isermann, Fault-Diagnosis Systems, Springer Berlin Heidelberg, Berlin, Heidelberg, 2006

    R. Isermann, Fault-Diagnosis Systems, Springer Berlin Heidelberg, Berlin, Heidelberg, 2006

    2006

  36. [45]

    S. G. Vrachimis, D. G. Eliades, R. Taormina, Z. Kapelan, A. Ostfeld, S. Liu, M. Kyri- akou, P. Pavlou, M. Qiu, M. M. Polycarpou, Battle of the leakage detection and isola- tion methods, Journal of Water Resources Planning and Management 148 (12) (2022) 04022068. doi:10.1061/(A...

    2022

  37. [46]

    Ashraf, J

    I. Ashraf, J. Strotherm, L. Hermes, B. Hammer, Physics-informed graph neural net- works for water distribution systems, in: M. J. Wooldridge, J. G. Dy, S. Natara- jan (Eds.), Thirty-Eighth AAAI Conference on Artificial Intelligence, AAAI 2024, Thirty-Sixth Conference on Innova...

    2024

  38. [47]

    Artelt, S

    A. Artelt, S. Vrachimis, D. Eliades, M. Polycarpou, B. Hammer, One explanation to rule them all – ensemble consistent explanations, in: R. Weber, O. Amir, T. Miller (Eds.), Workshop on XAI, 2022. doi:10.48550/arXiv.2205.08974. URLhttps://sites.google.com/view/xai2022

    2022 doi arXiv

  39. [48]

    Casillas, V

    M. Casillas, V. Puig, L. Garza-Castañón, A. Rosich, Optimal sensor placement for leak location in water distribution networks using genetic algorithms, Sensors 13 (11) (2013) 14984–15005. doi:10.3390/s131114984. 26

    2013 doi

  40. [635]

    doi:10.1061/(ASCE)0733-9429(2008)134:5(626)

    2008 doi

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.