arxiv: 2605.10847 · v1 · submitted 2026-05-11 · 💻 cs.LG

Recognition: 2 theorem links

· Lean Theorem

Conditional anomaly detection methods for patient-management alert systems

Amy Seybert, Gregory Cooper, Melissa Saul, Michal Valko, Milo\v{s} Hauskrecht, Shyam Visweswaran

Authors on Pith no claims yet

Pith reviewed 2026-05-12 04:59 UTC · model grok-4.3

classification 💻 cs.LG

keywords conditional anomaly detectioninstance-based methodsdistance metricsmetric learningmedical alert systemspneumonia admission decisionsheparin induced thrombocytopeniapatient management

0 comments

The pith

Instance-based methods with optimized distance metrics detect conditional anomalies in medical patient decisions.

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

The paper investigates instance-based approaches to conditional anomaly detection, where anomalies in one set of attributes are identified relative to the values of other conditioning attributes. It tests various distance metrics and metric learning techniques to improve how these methods select the most relevant examples from the data. The authors apply the methods to two real medical detection tasks involving unusual patient admissions for community-acquired pneumonia and atypical orders for an HPF4 test used to diagnose heparin-induced thrombocytopenia. A sympathetic reader would care because such methods could support more targeted alert systems in hospitals by focusing on context-dependent unusual patterns rather than global outliers.

Core claim

Instance-based conditional anomaly detection relies on a distance metric to locate the most critical examples in a dataset for deciding whether a given pattern is anomalous when conditioned on the remaining attributes. The work evaluates multiple metrics and metric learning procedures to tune this selection process and demonstrates performance gains on two clinical problems: flagging atypical admission decisions for patients with community-acquired pneumonia and identifying unusual orders of the HPF4 test for confirming heparin-induced thrombocytopenia.

What carries the argument

Instance-based conditional anomaly detection that uses a distance metric to retrieve and compare the most relevant conditioning examples from the dataset.

If this is right

The methods produce more accurate flags for unusual admission decisions in pneumonia cases.
They similarly improve detection of atypical HPF4 test orders that may indicate heparin-induced thrombocytopenia.
Metric learning can be used to adapt the distance function to the specific structure of each clinical dataset.
Instance-based selection of conditioning examples outperforms non-optimized alternatives for these conditional tasks.

Where Pith is reading between the lines

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

The same metric-optimization approach might transfer to other hospital alert tasks such as medication dosing or lab result interpretation.
Integration into electronic health record systems could reduce unnecessary alerts by conditioning on patient context.
Future work could test whether the learned metrics remain stable when the underlying patient population changes over time.

Load-bearing premise

Suitable distance metrics and metric learning procedures can be found that reliably separate conditional anomalies from normal patterns in these medical datasets without further unstated adjustments.

What would settle it

A direct comparison on the pneumonia admission and HPF4 ordering datasets showing that no tested distance metric or metric learning variant improves detection performance over a simple unconditional anomaly detector would disprove the claimed benefit.

Figures

Figures reproduced from arXiv: 2605.10847 by Amy Seybert, Gregory Cooper, Melissa Saul, Michal Valko, Milo\v{s} Hauskrecht, Shyam Visweswaran.

**Figure 1.** Figure 1: ROC curve for the anomaly detection method based on the and SVM projections on the HIT dataset. The methods are compared to the sensitivity and specificity of the rule based detector. total of 274 HPF4 orders were associated with these states (prior of a test order is 0.79%) Each data–point generated consisted of a total of 43 features that included recent platelets, platelet trends, platelet drops from n… view at source ↗

read the original abstract

Anomaly detection methods can be very useful in identifying unusual or interesting patterns in data. A recently proposed conditional anomaly detection framework extends anomaly detection to the problem of identifying anomalous patterns on a subset of attributes in the data. The anomaly always depends (is conditioned) on the value of remaining attributes. The work presented in this paper focuses on instance-based methods for detecting conditional anomalies. The methods rely on the distance metric to identify examples in the dataset that are most critical for detecting the anomaly. We investigate various metrics and metric learning methods to optimize the performance of the instance-based anomaly detection methods. We show the benefits of the instance-based methods on two real-world detection problems: detection of unusual admission decisions for patients with the community-acquired pneumonia and detection of unusual orders of an HPF4 test that is used to confirm Heparin induced thrombocytopenia - a life-threatening condition caused by the Heparin therapy.

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.

Desk Editor's Note private letter to a colleague

This applies instance-based distance methods plus metric learning to an existing conditional anomaly detection setup on two medical alert tasks, but the gains rest on unverified metric choices and lack visible quantitative backing.

read the letter

The core of this paper is taking a conditional anomaly detection framework and testing instance-based approaches that use distance metrics to pull out the most relevant examples for spotting anomalies in subsets of attributes. They try various metrics and some metric learning to improve detection on two real medical problems: unusual admission decisions for community-acquired pneumonia patients and odd HPF4 test orders tied to heparin-induced thrombocytopenia risk. That focus on practical alert systems is the main draw, and applying the methods to these high-stakes cases gives a concrete sense of where conditional anomalies could matter in patient management. They do a reasonable job laying out why instance-based methods fit here, since the anomaly depends on the rest of the patient profile. The exploration of different distances and learning options shows they thought about how to make the approach work on mixed medical data. That said, the abstract gives no numbers, no baselines, no error bars, and no clear description of how the metrics were chosen or validated on held-out data. Medical records often mix categorical and numeric fields with missing values, so standard distances can fall apart without careful adaptation. If the metric optimization was done in a dataset-specific way without independent checks, the reported benefits could be fragile rather than a general property of the instance-based route. The stress-test concern about robustness lands because nothing in the provided text shows that the separation of anomalous cases holds up under different metric choices or data splits. This is the kind of work that might interest people building anomaly tools for clinical decision support, especially if they already know the conditional framework. A reader looking for new theory or first-principles results will find little, but someone wanting examples of metric tuning in healthcare data could pick up ideas. The paper deserves a serious referee because the application area is useful and the methods are straightforward enough to evaluate once the full experiments are in front of reviewers. I would not cite it as is, but it could be worth a look after revisions that add proper comparisons and validation details.

Referee Report

2 major / 0 minor

Summary. The manuscript introduces instance-based methods for conditional anomaly detection that rely on distance metrics to surface critical examples for identifying anomalies on a subset of attributes conditioned on the rest. It examines various metrics and metric learning approaches to optimize performance and claims to demonstrate benefits on two real-world medical tasks: detecting unusual admission decisions for community-acquired pneumonia patients and unusual HPF4 test orders for confirming heparin-induced thrombocytopenia.

Significance. If the empirical claims hold with proper validation, the work could support more interpretable anomaly detection in clinical alert systems. The focus on instance-based conditional methods addresses a relevant gap in applying anomaly detection to conditioned medical decisions. However, the absence of any quantitative results, baselines, or robustness checks in the manuscript substantially reduces its potential impact and verifiability.

major comments (2)

[Abstract] Abstract: the claim that instance-based methods deliver benefits on the two medical detection problems is unsupported because the text supplies no quantitative results, error bars, baseline comparisons, or method details, preventing any assessment of whether the distance metrics actually separate conditional anomalies.
The central claim depends on distance metric optimization or learning to reliably identify conditional anomalies, yet no description is given of the metric selection procedure, handling of mixed categorical/numeric features and missing values, or held-out validation for the metric itself; without this, the reported benefits cannot be distinguished from dataset-specific tuning.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback, which identifies key gaps in the presentation of our empirical results and methodological details. We will revise the manuscript to address these points directly.

read point-by-point responses

Referee: [Abstract] Abstract: the claim that instance-based methods deliver benefits on the two medical detection problems is unsupported because the text supplies no quantitative results, error bars, baseline comparisons, or method details, preventing any assessment of whether the distance metrics actually separate conditional anomalies.

Authors: We agree that the abstract as written does not include quantitative results, error bars, baseline comparisons, or sufficient method details, making it impossible to evaluate the claims from the abstract alone. This is a valid criticism of the submitted version. In the revision, we will update the abstract to summarize the key quantitative findings from our experiments on the two medical datasets, including performance metrics with error bars and comparisons to baselines, so that the benefits of the instance-based methods and optimized distance metrics can be assessed. revision: yes
Referee: The central claim depends on distance metric optimization or learning to reliably identify conditional anomalies, yet no description is given of the metric selection procedure, handling of mixed categorical/numeric features and missing values, or held-out validation for the metric itself; without this, the reported benefits cannot be distinguished from dataset-specific tuning.

Authors: We acknowledge that the manuscript provides insufficient detail on the metric selection and optimization procedure, the handling of mixed categorical/numeric features and missing values, and the use of held-out validation for the metrics. These omissions make it difficult to rule out dataset-specific tuning. In the revised manuscript, we will add a detailed methods subsection covering the metric selection process (including any learning algorithms), feature handling strategies, missing value treatment, and the held-out validation protocol used to tune and evaluate the metrics independently of the final anomaly detection results. This will strengthen the evidence that the reported benefits are robust. revision: yes

Circularity Check

0 steps flagged

No significant circularity in empirical demonstration on external medical data

full rationale

The paper presents instance-based conditional anomaly detection methods that rely on distance metrics and metric learning. It demonstrates performance benefits empirically on two external real-world medical datasets (unusual admission decisions for community-acquired pneumonia patients and unusual HPF4 test orders for heparin-induced thrombocytopenia). No equations, derivations, or mathematical claims are present that could reduce to self-definition, fitted inputs renamed as predictions, or self-citation chains. The work is self-contained against external benchmarks with no load-bearing internal reductions.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract mentions no free parameters, mathematical axioms, or new invented entities; the approach relies on standard distance metrics and metric learning without further specification.

pith-pipeline@v0.9.0 · 5462 in / 1013 out tokens · 46164 ms · 2026-05-12T04:59:16.081845+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear
We propose an anomaly detection framework that solves the problem by building a measure d(.) that reflects the severity with which an example differs from conditional (context-to-target) patterns... exploits discriminant functions... support vector machines

Reference graph

Works this paper leans on

300 extracted references · 300 canonical work pages

[1]

Roweis and Geoffrey E

Jacob Goldberger and Sam T. Roweis and Geoffrey E. Hinton and Ruslan Salakhutdinov , title =. NIPS , year =

work page
[2]

Xing and Michael I

Eric P. Xing and Michael I. Jordan and Stuart J. Russell , title =. Proceedings of UAI , year =

work page
[3]

and Han, Jiawei and Wang, Jianyong and Yu, Philip S

Aggarwal, Charu C. and Han, Jiawei and Wang, Jianyong and Yu, Philip S. , title =. Proceedings of the 29th international conference on Very large data bases - Volume 29 , year =

work page
[4]

Aggarwal and Philip S

Charu C. Aggarwal and Philip S. Yu , title =. SIGMOD '01: Proceedings of the 2001 ACM SIGMOD international conference on Management of data , year =. doi:http://doi.acm.org/10.1145/375663.375668 , isbn =

work page doi:10.1145/375663.375668 2001
[5]

Aha and Dennis Kibler and Marc K

David W. Aha and Dennis Kibler and Marc K. Albert , title =. Mach. Learn. , year =. doi:http://dx.doi.org/10.1023/A:1022689900470 , issn =

work page doi:10.1023/a:1022689900470
[6]

Advances in Knowledge Discovery and Data Mining, 14th Pacific-Asia Conference, PAKDD 2010, Hyderabad, India, June 21-24, 2010

Leman Akoglu and Mary McGlohon and Christos Faloutsos , title =. Advances in Knowledge Discovery and Data Mining, 14th Pacific-Asia Conference, PAKDD 2010, Hyderabad, India, June 21-24, 2010. Proceedings. Part II , year =

work page 2010
[7]

Advances in Neural Information Processing Systems 18 , publisher =

Yasemin Altun and David McAllester and Mikhail Belkin , title =. Advances in Neural Information Processing Systems 18 , publisher =. 2006 , editor =

work page 2006
[8]

Machine Learning , year =

Christophe Andrieu and Nando de Freitas and Arnaud Doucet and Michael Jordan , title =. Machine Learning , year =

work page
[9]

Computer Vision and Image Understanding , year =

Ognjen Arandjelovic and Roberto Cipolla , title =. Computer Vision and Image Understanding , year =

work page
[10]

and Clawson, James R

Ashbrook, Daniel L. and Clawson, James R. and Lyons, Kent and Starner, Thad E. and Patel, Nirmal , title =. Proceeding of the twenty-sixth annual SIGCHI conference on Human factors in computing systems , year =

work page
[11]

Journal of Mathematical Analysis and Applications , year =

Karl Astrom , title =. Journal of Mathematical Analysis and Applications , year =

work page
[12]

Frank and A

A. Frank and A. Asuncion. UCI ML Repository. 2010

work page 2010
[13]

Gambling in a Rigged Casino: The Adversarial Multi-Armed Bandit problem , booktitle =

Peter Auer and Nicol. Gambling in a Rigged Casino: The Adversarial Multi-Armed Bandit problem , booktitle =. 1995 , pages =

work page 1995
[14]

Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition , year =

Boris Babenko and Ming-Hsuan Yang and Serge Belongie , title =. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition , year =

work page
[15]

ICML 2005 Workshop on Learning with Partially Classified Training Data , year =

Maria-Florina Balcan and Avrim Blum and Patrick Pakyan Choi and John Lafferty and Brian Pantano and Mugizi Robert Rwebangira and Xiaojin Zhu , title =. ICML 2005 Workshop on Learning with Partially Classified Training Data , year =

work page 2005
[16]

Journal of Machine Learning Research , year =

Aharon Bar-Hillel and Tomer Hertz and Noam Shental and Daphna Weinshall , title =. Journal of Machine Learning Research , year =

work page
[17]

IEEE Transactions on Systems, Man, and Cybernetics , year =

Andrew Barto and Richard Sutton and Charles Anderson , title =. IEEE Transactions on Systems, Man, and Cybernetics , year =

work page
[18]

Bates and Atul

David W. Bates and Atul. A. Gawande , title =. New England Journal of Medicine , year =

work page
[19]

J Am Med Inform Assoc , year =

David W Bates and Gilad J Kuperman and Samuel Wang and Tejal Gandhi and Anne Kittler and Lynn Volk and Cynthia Spurr and Ramin Khorasani and Milenko Tanasijevic and Blackford Middleton , title =. J Am Med Inform Assoc , year =. doi:10.1197/jamia.M1370 , institution =

work page doi:10.1197/jamia.m1370
[20]

Journal of Artificial Intelligence Research , year =

Jonathan Baxter and Peter Bartlett , title =. Journal of Artificial Intelligence Research , year =

work page
[21]

Journal of Artificial Intelligence Research , year =

Jonathan Baxter and Peter Bartlett and Lex Weaver , title =. Journal of Artificial Intelligence Research , year =

work page
[22]

Proceeding of the 17th Annual Conference on Learning Theory , year =

Mikhail Belkin and Irina Matveeva and Partha Niyogi , title =. Proceeding of the 17th Annual Conference on Learning Theory , year =

work page
[23]

Journal of Machine Learning Research , year =

Mikhail Belkin and Partha Niyogi and Vikas Sindhwani , title =. Journal of Machine Learning Research , year =

work page
[24]

1957 , author =

Dynamic Programming , publisher =. 1957 , author =

work page 1957
[25]

Mathematics of Computation , year =

Richard Bellman and Robert Kalaba and Bella Kotkin , title =. Mathematics of Computation , year =

work page
[26]

IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems , year =

Luca Benini and Alessandro Bogliolo and Giuseppe Paleologo and Giovanni De Micheli , title =. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems , year =

work page
[27]

Advances in Neural Information Processing Systems 11 , year =

Kristin Bennett and Ayhan Demiriz , title =. Advances in Neural Information Processing Systems 11 , year =

work page
[28]

1999 , author =

Nonlinear Programming , publisher =. 1999 , author =

work page 1999
[29]

Neural Computation , year =

Dimitri Bertsekas , title =. Neural Computation , year =

work page
[30]

1995 , author =

Dynamic Programming and Optimal Control , publisher =. 1995 , author =

work page 1995
[31]

1996 , author =

Neuro-Dynamic Programming , publisher =. 1996 , author =

work page 1996
[32]

1997 , author =

Introduction to Linear Optimization , publisher =. 1997 , author =

work page 1997
[33]

Mooney , title =

Mikhail Bilenko and Sugato Basu and Raymond J. Mooney , title =. ICML '04: Proceedings of the twenty-first international conference on Machine learning , year =. doi:http://doi.acm.org/10.1145/1015330.1015360 , isbn =

work page doi:10.1145/1015330.1015360
[34]

Blei and Michael I

David M. Blei and Michael I. Jordan , title =. Bayesian Analysis , year =

work page
[35]

Blei and Andrew Y

David M. Blei and Andrew Y. Ng and Michael I. Jordan , title =. J. Mach. Learn. Res. , year =

work page
[36]

Online Algorithms , year =

Avrim Blum , title =. Online Algorithms , year =

work page
[37]

ICML '01: Proceedings of the Eighteenth International Conference on Machine Learning , year =

Blum,, Avrim and Chawla,, Shuchi , title =. ICML '01: Proceedings of the Eighteenth International Conference on Machine Learning , year =

work page
[38]

and Hand, David J

Bolton, Richard J. and Hand, David J. , title =. Stat. Sci. , year =. doi:doi:10.1214/ss/1042727940 , keywords =

work page doi:10.1214/ss/1042727940
[39]

Journal of Machine Learning Research , year =

Olivier Bousquet and Andre Elisseeff , title =. Journal of Machine Learning Research , year =

work page
[40]

Journal of Artificial Intelligence Research , year =

Craig Boutilier and Thomas Dean and Steve Hanks , title =. Journal of Artificial Intelligence Research , year =

work page
[41]

Proceedings of the 13th International Conference on Machine Learning , year =

Craig Boutilier and Richard Dearden , title =. Proceedings of the 13th International Conference on Machine Learning , year =

work page
[42]

Exploiting Structure in Policy Construction , booktitle =

Craig Boutilier and Richard Dearden and Mois\'. Exploiting Structure in Policy Construction , booktitle =. 1995 , pages =

work page 1995
[43]

Advances in Neural Information Processing Systems 13 , year =

Justin Boyan and Michael Littman , title =. Advances in Neural Information Processing Systems 13 , year =

work page
[44]

Advances in Neural Information Processing Systems 7 , year =

Justin Boyan and Andrew Moore , title =. Advances in Neural Information Processing Systems 7 , year =

work page
[45]

Journal of Machine Learning Research , year =

Ronen Brafman and Moshe Tennenholtz , title =. Journal of Machine Learning Research , year =

work page
[46]

Proceedings of the 18th Conference on Uncertainty in Artificial Intelligence , year =

John Bresina and Richard Dearden and Nicolas Meuleau and Sailesh Ramakrishnan and David Smith and Rich Washington , title =. Proceedings of the 18th Conference on Uncertainty in Artificial Intelligence , year =

work page
[47]

and Kriegel, Hans-Peter and Ng, Raymond T

Breunig, Markus M. and Kriegel, Hans-Peter and Ng, Raymond T. and Sander, J\". LOF: identifying density-based local outliers , journal =. 2000 , volume =. doi:http://doi.acm.org/10.1145/335191.335388 , issn =

work page doi:10.1145/335191.335388 2000
[48]

Buntine , title =

W. Buntine , title =. IEEE Transactions on Knowledge and Data Engineering , year =

work page
[49]

Christopher J. C. Burges , title =. Data Mining and Knowledge Discovery , year =

work page
[50]

Bennett , title =

Colin Campbell and Kristin P. Bennett , title =. Advances in Neural Information Processing Systems 13, Papers from Neural Information Processing Systems (NIPS) 2000 , year =

work page 2000
[51]

Biometrika , year =

George Casella and Christian Robert , title =. Biometrika , year =

work page
[52]

2006 , author =

Prediction, Learning, and Games , publisher =. 2006 , author =

work page 2006
[53]

Chih-Chung Chang and Chih-Jen Lin , year =

work page
[54]

ICML '04: Proceedings of the twenty-first international conference on Machine learning , year =

Hong Chang and Dit-Yan Yeung , title =. ICML '04: Proceedings of the twenty-first international conference on Machine learning , year =. doi:http://doi.acm.org/10.1145/1015330.1015391 , isbn =

work page doi:10.1145/1015330.1015391
[55]

Inferring Identity Using Accelerometers in Television Remote Controls , booktitle =

Keng. Inferring Identity Using Accelerometers in Television Remote Controls , booktitle =. 2009 , pages =

work page 2009
[56]

Chapman and John N

Wendy W. Chapman and John N. Dowling and Gregory F. Cooper and Milos Hauskrecht and Michal Valko , title =. 2006 , abstract =

work page 2006
[57]

Proceedings of the 29th Annual ACM Symposium on Theory of Computing , year =

Moses Charikar and Chandra Chekuri and Tomas Feder and Rajeev Motwani , title =. Proceedings of the 29th Annual ACM Symposium on Theory of Computing , year =

work page
[58]

Eugene Charniak , title =. AI Mag. , year =

work page
[59]

Chawla and Aleksandar Lazarevic and Lawrence O

Nitesh V. Chawla and Aleksandar Lazarevic and Lawrence O. Hall and Kevin W. Bowyer , title =. PKDD , year =

work page
[60]

, title =

Chickering, David M. , title =. Learning from Data: Artificial Intelligence and Statistics V , publisher =. 1996 , editor =

work page 1996
[61]

IEEE Transactions on Automatic Control , year =

Chee-Seng Chow and John Tsitsiklis , title =. IEEE Transactions on Automatic Control , year =

work page
[62]

Proceedings of the 1999 IEEE / ACM International Conference on Computer-Aided Design , year =

Eui-Young Chung and Luca Benini and Giovanni de Micheli , title =. Proceedings of the 1999 IEEE / ACM International Conference on Computer-Aided Design , year =

work page 1999
[63]

1997 , author =

Spectral Graph Theory , publisher =. 1997 , author =

work page 1997
[64]

Proceedings of the Workshop on Uncertainty in Artificial Intelligence , year =

Gregory Cooper , title =. Proceedings of the Workshop on Uncertainty in Artificial Intelligence , year =

work page
[65]

Cooper, G. F. and Herskovits, E. , title =. Machine Learning , year =

work page
[66]

Proceedings of the 25th International Conference on Machine Learning , year =

Corinna Cortes and Mehryar Mohri and Dmitry Pechyony and Ashish Rastogi , title =. Proceedings of the 25th International Conference on Machine Learning , year =

work page
[67]

Crammer,, Koby and Singer,, Yoram , title =. J. Mach. Learn. Res. , year =

work page
[68]

Advances in Neural Information Processing Systems 8 , year =

Robert Crites and Andrew Barto , title =. Advances in Neural Information Processing Systems 8 , year =

work page
[69]

Mathematical Programming , year =

James Daniel , title =. Mathematical Programming , year =

work page
[70]

2009 , owner =

Kaustav Das , title =. 2009 , owner =

work page 2009
[71]

Cooper , title =

Denver Dash and Gregory F. Cooper , title =. ICML '02: Proceedings of the Nineteenth International Conference on Machine Learning , year =

work page
[72]

Davis and Brian Kulis and Prateek Jain and Suvrit Sra and Inderjit S

Jason V. Davis and Brian Kulis and Prateek Jain and Suvrit Sra and Inderjit S. Dhillon , title =. ICML '07: Proceedings of the 24th international conference on Machine learning , year =. doi:http://doi.acm.org/10.1145/1273496.1273523 , isbn =

work page doi:10.1145/1273496.1273523
[73]

Machine Learning , year =

Peter Dayan and Terry Sejnowski , title =. Machine Learning , year =

work page
[74]

Computational Intelligence , year =

Thomas Dean and Keiji Kanazawa , title =. Computational Intelligence , year =

work page
[75]

Proceedings of the 15th Conference on Uncertainty in Artificial Intelligence , year =

Richard Dearden and Nir Friedman and David Andre , title =. Proceedings of the 15th Conference on Uncertainty in Artificial Intelligence , year =

work page
[76]

Proceedings of the 15th International Joint Conference on Artificial Intelligence , year =

Rina Dechter , title =. Proceedings of the 15th International Joint Conference on Artificial Intelligence , year =

work page
[77]

Proceedings of the 12th Conference on Uncertainty in Artificial Intelligence , year =

Rina Dechter , title =. Proceedings of the 12th Conference on Uncertainty in Artificial Intelligence , year =

work page
[78]

Proceedings of the 2006 IEEE / ACM International Conference on Computer-Aided Design , year =

Gaurav Dhiman and Tajana Simunic , title =. Proceedings of the 2006 IEEE / ACM International Conference on Computer-Aided Design , year =

work page 2006
[79]

Pazzani , title =

Pedro Domingos and Michael J. Pazzani , title =. Machine Learning , year =

work page
[80]

Doyle and J

Doyle, Peter G. and Snell, Laurie J. , title =. 2000 , abstract =. math/0001057 , keywords =

work page arXiv 2000

Showing first 80 references.