Pith. sign in

REVIEW 4 major objections 6 minor 79 references

Integrating Probabilistic Trees and Causal Networks for Clinical and Epidemiological Data

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read PCF fuses causal Bayesian networks with probability trees to predict outcomes, quantify factor impacts, and simulate hypothetical interventions, with accuracy comparable to standard machine-learning models.

desk verdict PCF is a plausible incremental pipeline with credible predictive results, but its causal intervention and counterfactual claims do not follow from the method as described. read the letter →

arxiv 2501.15973 v1 pith:GGZ6YKNY submitted 2025-01-27 cs.LG q-bio.QM

Sheresh Zahoor , Pietro Liò , Gaël Dias , Mohammed Hasanuzzaman This is my paper · ORCID
classification cs.LGq-bio.QM
keywords causalBayesiannetworksprobabilitytreesinferenceinterventionalanalysiscounterfactualhealthcaremachinelearningSHAPensemble
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

PCF is a proposed framework that joins Causal Bayesian Networks with ensembles of Probability Trees so that a single model can predict a clinical outcome, quantify how strongly each factor influences it, and simulate what would happen under a hypothetical intervention. The CBN supplies a causal ordering of variables; the trees express the resulting conditional probabilities as branches; interventions are represented by forcing a branch to probability 1 and reading the outcome as $P(A \mid \mathrm{do}(B))$. On the MIMIC-IV, Framingham Heart Study, and BRFSS-2015 diabetes datasets, the framework reports predictive accuracy and AUC comparable to standard classifiers such as gradient boosting, random forests, and logistic regression, while also producing sensitivity analyses and SHAP explanations of factor importance. The paper's point is that prediction and causal reasoning need not be separate toolkits: one tree-based model can give clinicians both a risk score and an answer to 'what if we changed this factor?'

What carries the argument

The load-bearing object is the Probability Tree, a branching structure in which nodes are events and edges carry conditional probabilities, whose branching order is fixed by the topological order of a model-averaged Causal Bayesian Network. The CBN contributes the causal parent–child relationships that decide which variables branch before which, while empirical data determine the transition probabilities at each branch. Interventions are implemented by setting a chosen transition probability to 1, giving the paper's $P(A \mid \mathrm{do}(B))$ semantics; counterfactuals are implemented by altering the variable order or resetting factual states in a modified tree. The ensemble averages predictions over $k$ trees built on disjoint data subsets, and the averaged prediction is thresholded to produce a class label.

What would settle it

Compare the model's $P(\mathrm{CHD}=1 \mid \mathrm{do}(\mathrm{cigsPerDay}=0))$ estimate for heavy smokers against the observed 10-year coronary heart disease incidence among smokers who quit during the Framingham follow-up; if the model's counterfactual probability does not fall within the cohort's confidence interval, the tree-clamping operation is not reproducing a real intervention effect.

Watch

Extended reading notes

Core claim

The authors claim that the causal structure of a Causal Bayesian Network can fix the variable order of a Probability Tree, turning the tree into a reasoning tool rather than just a predictor. They learn several network structures from the data, average them into a single DAG (adding or reversing edges to avoid cycles), take its topological order, and build an ensemble of trees on random subsets of the data using that order. Transition probabilities are empirical counts of a variable given its CBN parent, and the root split reflects the marginal distribution of the target. An intervention is performed by setting a branch's transition probability to 1, which the paper writes as $P(A \mid \mathrm{do}(B))$, and counterfactuals are obtained by reordering variables or replacing observed states with hypothetical ones and recomputing path probabilities. On MIMIC-IV, Framingham, and BRFSS-2015, the method achieves accuracy and AUC-ROC within the range of the compared machine-learning baselines while adding sensitivity analysis, SHAP feature attribution, and the ability to simulate interventions.

Load-bearing premise

The causal claims stand on the assumption that the network learned from observational data is the true causal graph with all confounders observed, and that forcing a branch in the fitted tree to probability 1 is the same as intervening in the world.

Editorial extensions

If this is right

  • A clinician could query the model for the effect of lowering blood pressure, quitting smoking, or reducing BMI on the outcome and receive a probability change computed from the same model that gives the risk prediction.
  • The CBN-derived variable order removes the need for an expert to hand-specify the probability tree's branching sequence, replacing a subjective step with a data-driven one.
  • On the three datasets, the framework's accuracy and AUC-ROC are comparable to gradient boosting, random forests, support vector machines, and logistic regression, with a more balanced sensitivity–specificity profile than several baselines.
  • Sensitivity analysis and SHAP values can be produced from the same model, letting clinicians see both parameter-level and feature-level explanations.
  • Pre-trained PCF models with their learned causal graphs could be shared across hospitals, so smaller institutions can start from collective causal knowledge and refine it with local data.

Reading between the lines

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

  • Because each node's transition probability is conditioned on a single parent, the tree factorization may not capture the full joint distribution implied by the CBN, so an intervention in one branch could propagate differently than the network's multi-parent structure would dictate.
  • A direct way to test the causal layer would be to benchmark PCF's intervention curves against a known randomized effect, such as the Framingham cohort's observed outcomes following smoking cessation, and check whether the model's counterfactual reductions match the longitudinal data.
  • The same tree-building recipe could be applied to continuous outcomes, time-to-event data, or multi-category outcomes, and the intervention operation could be adapted to shift transition probabilities fractionally instead of forcing them to 1.
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 the Probabilistic Causal Fusion (PCF) framework, which combines Causal Bayesian Network structure learning (via Bayesys model averaging) with an ensemble of probability trees whose variable ordering is taken from a topological sort of the learned graph. PCF is evaluated on three healthcare datasets (MIMIC-IV ICU length of stay, Framingham CHD risk, and BRFSS diabetes risk) and compared against conventional ML classifiers. The authors report comparable predictive accuracy/AUC and use the learned structure for sensitivity analysis, SHAP explanations, 'interventions' implemented by forcing PTree transition probabilities to 1, and 'counterfactual' statements following the probability-tree algorithms of Genewein et al. [8]. The central claim is that PCF provides a single tool that quantifies factor impacts and simulates hypothetical interventions from observational data while maintaining predictive performance comparable to standard ML.

Significance. If the causal claims were established, the framework would be a useful contribution: the predictive component is compared with several baselines on three real datasets and the reported metrics (Tables 1-3) support the 'comparable predictive performance' part of the claim. The method description is sufficiently detailed to permit re-implementation, and the use of external ML baselines avoids circularity. However, the distinctive added value—causal intervention and counterfactual reasoning from observational data—is not supported by the current derivation or validation. The intervention operation is defined purely as a change in an empirical tree's transition probabilities, with no causal identifiability conditions, no proof that the tree's factorization equals the CBN's joint distribution, and no causal baseline or ground-truth comparison. Thus the paper's significance currently rests on its predictive results plus a set of descriptive conditional analyses; the causal reasoning capability advertised in the abstract remains unsubstantiated.

major comments (4)
  1. [Section 5.3 and Section 3.2.1]
  2. [Algorithm 1 and Section 3.2.1(b)]
  3. [Section 5.4.1 and 5.4.2]
  4. [Section 5.1]
minor comments (6)
  1. [Section 5.3]
  2. [Figures 5 and 6]
  3. [Algorithm 2]
  4. [Section 3.1.2]
  5. [Section 5.4.2]
  6. [General]

Circularity Check

1 steps flagged · score 4.0 of 10

Intervention effect is self-definitional: do(B) is defined as editing the tree's transition probabilities, while the predictive benchmarking core is non-circular.

  1. self definitional [Section 5.3, Intervention (page 10)]
    "In the context of PCF, intervention involves the strategic modification of transition probabilities to ensure a specific event occurs with certainty (probability of 1). This approach allows for the exploration of conditional probabilities represented as P(A|do(B)), indicating the probability of event A occurring given that eventB is enforced."

    The paper defines an intervention as the act of editing the PTree's transition probabilities and then labels the resulting quantity P(A|do(B)). The do-notation is attached to the tree's own modified conditional probability, with no independent causal identification: no condition is stated or tested that the learned CBN is the true causal graph, that all confounders are observed, or that the single-parent tree factorization equals the CBN joint. The 'intervention effect' therefore reduces by definition to the edited tree's output, and the causal semantics are supplied by notation rather than derived from the data or from an identified causal model.

full rationale

The predictive core is independent: PTree transition probabilities are empirical counts, and PCF's accuracy/AUC are benchmarked against external ML models (Tables 1-3), so no fitted parameter is renamed as a prediction. No load-bearing self-citation appears: [7] is background, and the counterfactual algorithm reference [8] is non-overlapping. The central causal layer, however, is self-definitional in one place: Section 5.3's intervention is literally the act of modifying transition probabilities, so the reported P(A|do(B)) is the tree's edited probability rather than a quantity derived from an identified causal model. The counterfactual section inherits this definition. The practice of choosing features from the literature and then checking agreement with the same literature is a soft-validation weakness but not a by-construction reduction, so it does not increase the score beyond 4.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The framework rests on assumptions that are not fully stated or tested. The main free parameters are unreported hyperparameters; the main axioms are the causal validity of the learned graph and the PTree intervention semantics; no new entities are introduced.

free parameters (6)
  • Ensemble size k (number of PTrees/data subsets)
    Section 3.2.2 splits D into k subsets but does not report k; this controls variance and predictive performance.
  • Pruning threshold
    Algorithm 1 requires a pruning threshold for retaining branches; the value is never specified, and it directly determines tree structure.
  • Classification threshold tau
    Section 3.2.3 uses tau to map averaged probabilities to classes; the default value and selection procedure are not reported, which affects accuracy versus sensitivity trade-off.
  • Model averaging edge frequency threshold
    Section 3.1.1 adds edges to Gavg when freq(e) > threshold; the threshold is not reported and changes the learned graph and hence the variable order.
  • Resampling method choice (SMOTE vs ADASYN)
    Section 5.2 says the superior method was chosen per dataset; this is a data-dependent modeling choice that affects all reported metrics.
  • MIMIC LOS threshold (4 days) = 4 days
    Section 4.1.1 categorizes LOS at the 75th percentile (Q3=4.0) of the dataset; the target definition depends on this data-derived cut, though it is clinically motivated.
assumptions (6)
  • domain assumption The DAG learned by score-based structure learning from observational data represents the true causal graph.
    Section 3.1.1 uses HC, TABU, SaiyanH, MAHC, GES and model averaging to find Gavg, then treats the topological order as a causal influence sequence; no causal sufficiency, faithfulness, or orientation identifiability conditions are stated.
  • domain assumption No unmeasured confounding and correct variable set.
    Intervention and counterfactual claims in Sections 5.3-5.4 require that all common causes are observed; the paper does not state or test this.
  • ad hoc to paper Setting a transition probability to 1 in a PTree corresponds to Pearl's do-operator.
    Section 5.3 defines intervention as strategic modification of transition probabilities and writes P(A|do(B)); this equivalence is asserted, not derived, and requires the tree to be a faithful representation of the causal generative process.
  • ad hoc to paper PTree transition probabilities conditioned on a single CBN parent reproduce the relevant joint distribution.
    Algorithm 1 uses P(Xcurrent|Xparent) = Count(...)/Count(...) with only one parent; along a path, this ignores other ancestors and evidence, so the tree's distribution may diverge from the empirical joint distribution.
  • ad hoc to paper The counterfactual algorithms of Genewein et al. [8] remain valid for the empirically constructed PCF trees.
    Section 5.4.2 does not provide the algorithmic details or conditions; it relies entirely on [8].
  • standard math Standard probability and SHAP mathematics.
    The prediction averaging, conditional probability, and SHAP formula (Section 3.2.4) assume standard probability theory and game-theoretic attribution.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Integrating Probabilistic Trees and Causal Networks for Clinical and Epidemiological Data." pith.science (2026). https://pith.science/paper/GGZ6YKNY

@misc{pith2026250115973,
  author       = {Pith},
  title        = {Pith review of: Integrating Probabilistic Trees and Causal Networks for Clinical and Epidemiological Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GGZ6YKNY}},
  note         = {Machine review of arXiv:2501.15973}
}
read the original abstract

Healthcare decision-making requires not only accurate predictions but also insights into how factors influence patient outcomes. While traditional Machine Learning (ML) models excel at predicting outcomes, such as identifying high risk patients, they are limited in addressing what-if questions about interventions. This study introduces the Probabilistic Causal Fusion (PCF) framework, which integrates Causal Bayesian Networks (CBNs) and Probability Trees (PTrees) to extend beyond predictions. PCF leverages causal relationships from CBNs to structure PTrees, enabling both the quantification of factor impacts and simulation of hypothetical interventions. PCF was validated on three real-world healthcare datasets i.e. MIMIC-IV, Framingham Heart Study, and Diabetes, chosen for their clinically diverse variables. It demonstrated predictive performance comparable to traditional ML models while providing additional causal reasoning capabilities. To enhance interpretability, PCF incorporates sensitivity analysis and SHapley Additive exPlanations (SHAP). Sensitivity analysis quantifies the influence of causal parameters on outcomes such as Length of Stay (LOS), Coronary Heart Disease (CHD), and Diabetes, while SHAP highlights the importance of individual features in predictive modeling. By combining causal reasoning with predictive modeling, PCF bridges the gap between clinical intuition and data-driven insights. Its ability to uncover relationships between modifiable factors and simulate hypothetical scenarios provides clinicians with a clearer understanding of causal pathways. This approach supports more informed, evidence-based decision-making, offering a robust framework for addressing complex questions in diverse healthcare settings.

Figures

Figures reproduced from arXiv: 2501.15973 by the authors.

Figure 1
Figure 1. Different steps involved in the PCF framework. The first module addresses data pre-processing to shape the input required for the CBN. The next module involves generating individual CBNs and creating a model-averaging graph. Subsequently, the ensemble of PTrees is developed based on the variable order from the model-averaging graph. The final module involves evaluating the overall performance of PCF. on one factor m… view at source ↗
Figure 2
Figure 2. Sensitivity Analysis for LOS, Diabetes, and Framingham datasets. [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗
Figure 3
Figure 3. SHAP plot showing feature impacts on predictions for LOS, CHD, and Diabetes. [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Probability change of los given interventions on heart [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Probability change of TenYearCHD given interventions on sysBP, diaBP, totChol, BMI, education, glucose, heartRate and cigsPerDay. [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Probability change of Diabetes given interventions on sysBP, diaBP, totChol, BMI, education, glucose, heartRate and cigsPerDay. [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: The process of developing and sharing pre-trained PCFs by large hospitals with extensive datasets. The feedback loop illustrates how re-ordering variables [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]
Figure 8
Figure 8. Figure 8: Visualisation of Counterfactual statements [PITH_FULL_IMAGE:figures/full_fig_p017_8.png]
Figure 9
Figure 9. Figure 9: Line plots showing the probability of ICU stay exceeding 4 days under factual and counterfactual scenarios for various health variables. [PITH_FULL_IMAGE:figures/full_fig_p017_9.png]
Figure 10
Figure 10. Figure 10: Probability distribution of TenYearCHD for factual and counterfactual scenarios for various variables. [PITH_FULL_IMAGE:figures/full_fig_p019_10.png]
Figure 11
Figure 11. Figure 11: Probability Distribution of Diabetes for factual and counterfactual scenarios for various variables. [PITH_FULL_IMAGE:figures/full_fig_p019_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

79 extracted references · 64 canonical work pages

  1. [8]

    Algorithms for Causal Reasoning in Probability Trees

    T. Genewein, T. McGrath, G. Del ´etang, V . Mikulik, M. Mar- tic, S. Legg, P. A. Ortega, Algorithms for causal reasoning in probability trees, arXiv preprint arXiv:2010.12237 (2020). doi:https://doi.org/10.48550/arXiv.2010.12237

    work page Pith review arXiv doi:10.48550/arxiv.2010.12237 2020

  2. [7]

    E. L. Ambags, G. Capitoli, V . L. Imperio, M. Proven- zano, M. S. Nobile, P. Li `o, Assisting clinical practice with fuzzy probabilistic decision trees (2023). arXiv:2304.07788, doi:https://doi.org/10.48550/arXiv.2304.07788

    work page Pith review arXiv doi:10.48550/arxiv.2304.07788 2023

  3. [1]

    Pearl, Causality: Models, Reasoning and Inference, Cambridge University Press, (2009)

    J. Pearl, Causality: Models, Reasoning and Inference, Cambridge University Press, (2009). URL http://bayes.cs.ucla.edu/BOOK-2K/neuberg-review.pdf

    work page 2009

  4. [2]

    Sanchez, J

    P. Sanchez, J. P. V oisey, T. Xia, H. I. Watson, A. Q. O’Neil, S. A. Tsaftaris, Causal machine learning for healthcare and preci- sion medicine, Royal Society Open Science 9 (8) (2022) 220638. doi:https://doi.org/10.1098/rsos.220638. URL https://royalsocietypublishing.org/doi/ 10.1098/rsos.220638

    doi:10.1098/rsos.220638 2022

  5. [3]

    Kopitar, P

    L. Kopitar, P. Kocbek, L. Cilar, A. Sheikh, G. Stiglic, Early detection of type 2 diabetes mellitus using machine learning-based prediction models, Scientific reports 10 (1) (2020) 11981

    work page 2020

  6. [4]

    Feuerriegel, D

    S. Feuerriegel, D. Frauen, V . Melnychuk, J. Schweisthal, K. Hess, A. Curth, S. Bauer, N. Kilbertus, I. S. Kohane, M. van der Schaar, Causal machine learning for predicting treatment outcomes, Nature Medicine 30 (4) (2024) 958–968

    work page 2024

  7. [5]

    Prosperi, Y

    M. Prosperi, Y . Guo, M. Sperrin, J. S. Koopman, J. S. Min, X. He, S. Rich, M. Wang, I. E. Buchan, J. Bian, Causal inference and counterfactual pre- diction in machine learning for actionable healthcare, Nature Machine Intelligence 2 (7) (2020) 369–375. doi:https: //doi.org/10.1038/s42256- 020-0197-y. URL https://www.nature.com/articles/s42256-020-0197-y

    doi:10.1038/s42256- 2020

  8. [6]

    V . K. Rajput, M. K. Kaltoft, J. Dowie, The causal plausibility deci- sion in healthcare, in: pHealth 2022, IOS Press, 2022, pp. 75–88. doi:10.3233/SHTI220965

    work page doi:10.3233/shti220965 2022

Show all 79 references
  1. [9]

    L. Chen, P. Ji, Y . Ma, Y . Rong, J. Ren, Custom machine learning al- gorithm for large-scale disease screening-taking heart disease data as an example, Artificial Intelligence in Medicine 146 (2023) 102688

    2023

  2. [10]

    T. M. Usman, Y . K. Saheed, A. Nsang, A. Ajibesin, S. Rakshit, A system- atic literature review of machine learning based risk prediction models for diabetic retinopathy progression, Artificial intelligence in medicine 143 (2023) 102617

    2023

  3. [11]

    Mennickent, A

    D. Mennickent, A. Rodr ´ıguez, M. Far ´ıas-Jofr´e, J. Araya, E. Guzm ´an- Guti´errez, Machine learning-based models for gestational diabetes mel- litus prediction before 24–28 weeks of pregnancy: A review, Artificial Intelligence in Medicine 132 (2022) 102378

    2022

  4. [12]

    M. M. Ahsan, Z. Siddique, Machine learning-based heart disease diagno- sis: A systematic literature review, Artificial Intelligence in Medicine 128 (2022) 102289

    2022

  5. [13]

    H. Ma, D. Li, J. Zhao, W. Li, J. Fu, C. Li, Hr-bgcn: Predicting readmission for heart failure from electronic health records, Artificial Intelligence in Medicine 150 (2024) 102829

    2024

  6. [14]

    L. Yao, Z. Chu, S. Li, Y . Li, J. Gao, A. Zhang, A survey on causal infer- ence, ACM Transactions on Knowledge Discovery from Data (TKDD) 15 (5) (2021) 1–46

    2021

  7. [15]

    C. J. Blumberg, Causal inference for statistics, social, and biomedical sciences: An introduction (2016)

    2016

  8. [16]

    Belthangady, S

    C. Belthangady, S. Giampanis, I. Jankovic, W. Stedden, P. Alves, S. Chong, C. Knott, B. Norgeot, Causal deep learning reveals the compar- ative effectiveness of antihyperglycemic treatments in poorly controlled diabetes, Nature Communications 13 (1) (2022) 6921

    2022

  9. [17]

    J. G. Richens, C. M. Lee, S. Johri, Improving the accuracy of medical diagnosis with causal machine learning, Nature communications 11 (1) (2020) 3923

    2020

  10. [18]

    P. W. Holland, Statistics and causal inference, Journal of the American statistical Association 81 (396) (1986) 945–960

    1986

  11. [19]

    Rajendran, M

    K. Rajendran, M. Jayabalan, V . Thiruchelvam, Predicting breast cancer via supervised machine learning methods on class imbalanced data, Inter- national Journal of Advanced Computer Science and Applications 11 (8) (2020)

    2020

  12. [20]

    Shahmirzalou, M

    P. Shahmirzalou, M. J. Khaledi, M. Khayamzadeh, A. Rasekhi, Survival analysis of recurrent breast cancer patients using mix bayesian network, Heliyon 9 (10) (2023)

    2023

  13. [21]

    Jang, S.-J

    B.-S. Jang, S.-J. Chun, H. S. Choi, J. H. Chang, K. H. Shin, et al., Es- timating the risk and benefit of radiation therapy in (y) pn1 stage breast cancer patients: A bayesian network model incorporating expert knowl- edge (krog 22–13), Computer Methods and Programs in Biomedi...

    2024

  14. [22]

    J. M. Ordov ´as, D. Rios-Insua, A. Santos-Lozano, A. Lucia, A. Torres, A. Kosgodagan, J. M. Camacho, A bayesian network model for predicting cardiovascular risk, Computer methods and programs in biomedicine 231 (2023) pp. 107405. doi:10.1016/j.cmpb.2023.107405

    2023

  15. [23]

    I. J. Dahabreh, K. Bibbins-Domingo, Causal inference about the e ffects of interventions from observational studies in medical journals, JAMA (2024)

    2024

  16. [24]

    Constantinou, The Bayesys user manual, http://bayesianai.eecs.qmul.ac.uk/bayesys/, [Online] Accessed 21 April 2023 (2019)

    A. Constantinou, The Bayesys user manual, http://bayesianai.eecs.qmul.ac.uk/bayesys/, [Online] Accessed 21 April 2023 (2019)

    2019

  17. [25]

    Heckerman, D

    D. Heckerman, D. Geiger, D. M. Chickering, Learn- ing bayesian networks: The combination of knowledge and statistical data, Machine learning 20 (1995) 197–243. doi:https://doi.org/10.1023/A:1022623210503

    1995 doi

  18. [26]

    R. R. Bouckaert, Properties of bayesian belief network learn- ing algorithms, in: Proceedings of the Tenth International Con- ference on Uncertainty in Artificial Intelligence, Morgan Kauf- mann Publishers Inc., San Francisco, CA, USA, 1994, p. 102–109. doi:https://doi.org/10....

    1994 doi

  19. [27]

    Bouckaert, Bayesian belief networks: from construction to inference, Ph.D

    R. Bouckaert, Bayesian belief networks: from construction to inference, Ph.D. thesis, University of Utrecht, Utrecht, Netherlands (1995). URL https://core.ac.uk/download/pdf/39700264.pdf

    1995

  20. [28]

    A. C. Constantinou, Learning bayesian networks that enable full prop- agation of evidence, IEEE Access V ol. 8 (2020) pp. 124845–124856. doi:10.1109/ACCESS.2020.3006472. URL https://ieeexplore.ieee.org/document/9136714

    2020

  21. [29]

    A. C. Constantinou, Y . Liu, N. K. Kitson, K. Chobtham, Z. Guo, Ef- fective and efficient structure learning with pruning and model averaging strategies, Int. J. Approx. Reasoning V ol. 151 (C) (2022) pp. 292–321. doi:https://doi.org/10.1016/j.ijar.2022.09.016

    2022 doi

  22. [30]

    D. M. Chickering, C. Meek, Finding optimal bayesian networks, in: Pro- ceedings of the Eighteenth Conference on Uncertainty in Artificial Intel- ligence, UAI’02, Morgan Kaufmann Publishers Inc., San Francisco, CA, USA, 2002, p. 94–102

    2002

  23. [31]

    A. C. Constantinou, Z. Guo, N. K. Kitson, The impact of prior knowledge on causal structure learning, Knowledge and Information Systems 65 (8) (2023) 3385–3434. doi:https://doi.org/10.1007/s10115-023-01858-x. URL https://link.springer.com/article/10.1007/ s10115-023-01858-x

    2023 doi

  24. [32]

    Kjærul ff, L

    U. Kjærul ff, L. C. Van Der Gaag, Making sensitivity analysis computa- tionally efficient, arXiv preprint arXiv:1301.3868 (2013)

    2013 arXiv

  25. [33]

    BayesFusion, GeNIe Modeler USER MANUAL, https://support.bayesfusion.com/docs/GeNIe.pdf, [Online] Accessed 21 April 2023 (2023)

    2023

  26. [34]

    Biecek, T

    P. Biecek, T. Burzykowski, Explanatory model analysis: explore, explain, and examine predictive models, Chapman and Hall/CRC, 2021

    2021

  27. [35]

    J. C. Marshall, L. Bosco, N. K. Adhikari, B. Connolly, J. V . Diaz, T. Dor- man, R. A. Fowler, G. Meyfroidt, S. Nakagawa, P. Pelosi, et al., What is an intensive care unit? a report of the task force of the world federation of societies of intensive and critical care medicine,...

    2017 doi

  28. [36]

    M. H. Weil, W. Tang, From intensive care to critical care medicine: a historical perspective, American journal of respiratory and critical care medicine 183 (11) (2011) 1451–1453. doi:10.1164 /rccm.201008- 1341OE

    2011

  29. [37]

    G. H. Robinson, L. E. Davis, R. P. Leifer, Prediction of hospital length of stay, Health services research 1 (3) (1966) 287

    1966

  30. [38]

    Stone, R

    K. Stone, R. Zwiggelaar, P. Jones, N. Mac Parthal ´ain, A system- atic review of the prediction of hospital length of stay: Towards a unified framework, PLOS Digital Health 1 (4) (2022) e0000017. doi:10.1371/journal.pdig.0000017. URL https://www.ncbi.nlm.nih.gov/pmc /articles/...

    2022 doi

  31. [39]

    Johnson, L

    A. Johnson, L. Bulgarelli, T. Pollard, S. Horng, L. A. Celi, R. Mark, Mimic-iv (version 2.2) (2023). doi:https://doi.org/10.13026/6mm1-ek67. URL https://physionet.org/content/mimiciv/2.2/

    2023 doi

  32. [40]

    Hempel, S

    L. Hempel, S. Sadeghi, T. Kirsten, Prediction of intensive care unit length of stay in the mimic-iv dataset, Applied Sciences 13 (12) (2023). doi:10.3390/app13126930. URL https://www.mdpi.com/2076-3417/13/12/6930

    2023 doi

  33. [41]

    WHO, Cardiovascular diseases (CVDs), https://www.who.int/news-room/fact-sheets /detail/cardiovascular-diseases-(cvds), [Online] Accessed 24-April-2024 (2019)

    2019

  34. [42]

    S. S. Mahmood, D. Levy, R. S. Vasan, T. J. Wang, The framingham heart study and the epidemiology of cardiovascular disease: a historical perspective, The lancet 383 (9921) (2014) 999–1008. doi:10.1016/S0140- 6736(13)61752-3. URL https://www.ncbi.nlm.nih.gov/pmc/articles /PMC4159698/

    2014 doi

  35. [43]

    M. K. Ali, K. I. Galaviz, M. B. Weber, K. V . Narayan, The global burden of diabetes, Textbook of diabetes (2017) 65– 83doi:https://doi.org/10.1002/9781118924853.ch5

    2017 doi

  36. [44]

    X. An, Y . Zhang, W. Sun, X. Kang, H. Ji, Y . Sun, L. Jiang, X. Zhao, Q. Gao, F. Lian, et al., Early e ffective intervention can significantly reduce all-cause mortality in prediabetic patients: a systematic review and meta-analysis based on high-quality clinical studies, Fron...

    2024

  37. [45]

    CDC - Behavioral Risk Factor Surveillance System (BRFSS), http://www.cdc.gov/brfss/index.html, [Online] Accessed 21 March 2022 (2015)

    2015

  38. [46]

    M. P. LaValley, Logistic regression, Circulation 117 (18) (2008) 2395–

    2008

  39. [47]

    Suthaharan, S

    S. Suthaharan, S. Suthaharan, Decision tree learning, Machine Learning Models and Algorithms for Big Data Classification: Thinking with Examples for E ffective Learning (2016) 237– 269doi:https://doi.org/10.1007/978-1-4899-7641-3 10

    2016 doi

  40. [48]

    S. J. Rigatti, Random forest, Journal of Insurance Medicine 47 (1) (2017) 31–39. doi:10.17849/insm-47-01-31-39.1. URL https://pubmed.ncbi.nlm.nih.gov/28836909/

    2017

  41. [49]

    V . N. Vapnik, The nature of statistical learning theory (1995)

    1995

  42. [50]

    Cover, P

    T. Cover, P. Hart, Nearest neighbor pattern classification, IEEE Transactions on Information Theory 13 (1) (1967) 21–27. doi:10.1109/TIT.1967.1053964. URL https://ieeexplore.ieee.org/document/1053964

    1967

  43. [51]

    Mucherino, P

    A. Mucherino, P. J. Papajorgji, P. M. Pardalos, k-Nearest Neigh- bor Classification, Springer New York, 2009, pp. 83–106. doi:https://doi.org/10.1007/978-0-387-88615-2 4

    2009 doi

  44. [52]

    Freund, R

    Y . Freund, R. E. Schapire, A decision-theoretic generaliza- tion of on-line learning and an application to boosting, Jour- nal of Computer and System Sciences 55 (1) (1997) 119–139. doi:https://doi.org/10.1006/jcss.1997.1504. URL https://www.sciencedirect.com/science /article...

    1997

  45. [53]

    Friedman, Greedy function approximation: A gradient boosting machine, The Annals of Statistics 29 (11 2000)

    J. Friedman, Greedy function approximation: A gradient boosting machine, The Annals of Statistics 29 (11 2000). doi:10.1214/aos/1013203451

    2000

  46. [54]

    T. Chen, C. Guestrin, Xgboost: A scalable tree boosting system, in: Pro- ceedings of the 22nd ACM SIGKDD International Conference on Knowl- edge Discovery and Data Mining, Association for Computing Machinery, 2016, p. 785–794. doi:10.1145/2939672.2939785. URL https://doi.org/1...

    2016

  47. [55]

    Freund, R

    Y . Freund, R. E. Schapire, et al., Experiments with a new boosting algo- rithm, in: ICML, V ol. 96, 1996, pp. 148–156. doi:https: //dl.acm.org/doi /10.5555/3091696.3091715

    1996

  48. [56]

    N. V . Chawla, K. W. Bowyer, L. O. Hall, W. P. Kegelmeyer, Smote: syn- thetic minority over-sampling technique, Journal of artificial intelligence research 16 (2002) 321–357. doi:https://doi.org/10.1613/jair.953

    2002 doi

  49. [57]

    H. He, Y . Bai, E. A. Garcia, S. Li, Adasyn: Adaptive syn- thetic sampling approach for imbalanced learning, in: 2008 IEEE international joint conference on neural networks (IEEE world congress on computational intelligence), IEEE, 2008, pp. 1322–1328. doi:10.1109/IJCNN.2008.4...

    2008

  50. [58]

    C. Luo, Z. Duan, Z. Xia, Q. Li, B. Wang, T. Zheng, D. Wang, D. Han, Minimum heart rate and mortality after cardiac surgery: ret- rospective analysis of the multi-parameter intelligent monitoring in in- tensive care (mimic-iii) database, Scientific Reports 13 (02 2023). doi:10....

    2023 doi

  51. [59]

    Weiner, W

    I. Weiner, W. Mitch, J. Sands, Urea and ammonia metabolism and the control of renal nitrogen excretion, Clinical journal of the American Society of Nephrology : CJASN 10 (2014) 1444–58. doi:10.2215/CJN.10311013. URL https://www.ncbi.nlm.nih.gov/pmc/articles /PMC4527031/

    2014 doi

  52. [60]

    Tonelli, F

    M. Tonelli, F. Sacks, M. Arnold, L. Moye, B. Davis, M. Pfe ffer, Rela- tion between red blood cell distribution width and cardiovascular event rate in people with coronary disease, Circulation 117 (2) (2008) 163–168. doi:https://doi.org/10.1161/CIRCULATIONAHA.107.727545. URL h...

    2008

  53. [61]

    De Rosa, M

    S. De Rosa, M. Greco, M. Rauseo, M. G. Annetta, The good, the bad, and the serum creatinine: exploring the e ffect of muscle mass and nutrition, Blood Purification 52 (9-10) (2023) 775–785

    2023

  54. [62]

    M. A. Puskarich, U. Nandi, B. G. Long, A. E. Jones, Association between persistent tachycardia and tachypnea and in-hospital mortality among non-hypotensive emergency department patients admitted to the hospi- tal, Clinical and Experimental Emergency Medicine 4 (2017) 2 – 9. d...

    2017 doi

  55. [63]

    B. A. Cunha, K. W. Shea, Fever in the intensive care unit, Infectious Disease Clinics 10 (1) (1996) 185–209

    1996

  56. [64]

    A. C. Flint, C. Conell, X. Ren, N. M. Banki, S. L. Chan, V . A. Rao, R. B. Melles, D. L. Bhatt, E ffect of systolic and diastolic blood pressure on cardiovascular outcomes, New England Journal of Medicine 381 (3) (2019) 243–251. doi:10.1056/NEJMoa1803180. URL https://www.nejm....

    2019 doi

  57. [65]

    Poznyak, L

    A. Poznyak, L. Litvinova, P. Poggio, V . Sukhorukov, A. Orekhov, Effect of glucose levels on cardiovascular risk, Cells 11 (2022) 3034. doi:10.3390/cells11193034. URL https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC9562876/

    2022 doi

  58. [66]

    S. A. Peters, Y . Singhateh, D. Mackay, R. R. Huxley, M. Woodward, Total cholesterol as a risk factor for coronary heart disease and stroke in women compared with men: A systematic review and meta-analysis, Atheroscle- rosis 248 (2016) 123–131

    2016

  59. [67]

    Dharod, E

    A. Dharod, E. Z. Soliman, F. Dawood, H. Chen, S. Shea, S. Nazar- ian, A. G. Bertoni, M. Investigators, et al., Association of asymptomatic bradycardia with incident cardiovascular disease and mortality: the multi- ethnic study of atherosclerosis (mesa), JAMA internal medicine ...

    2016

  60. [68]

    Gallucci, A

    G. Gallucci, A. Tartarone, R. Lerose, A. V . Lalinga, A. M. Capobianco, Cardiovascular risk of smoking and benefits of smoking cessation, Jour- nal of thoracic disease 12 (7) (2020) 3866. doi:10.21037/jtd.2020.02.47. URL https://www.ncbi.nlm.nih.gov/pmc/articles /PMC7399440/

    2020 doi

  61. [69]

    Tillmann, J

    T. Tillmann, J. Vaucher, A. Okbay, H. Pikhart, A. Peasey, R. Kubinova, A. Pajak, S. Malyutina, F. Hartwig, K. Fischer, G. Veronesi, T. Palmer, 21 J. Bowden, G. Smith, M. Bobak, M. Holmes, Education and coronary heart disease: Mendelian randomisation study, BMJ 358 (2017) j3542...

    2017 doi

  62. [70]

    C. Held, N. Hadziosmanovic, P. Aylward, E. Hagstr ¨om, J. Hochman, R. Stewart, H. White, L. Wallentin, Body mass index and association with cardiovascular outcomes in patients with stable coronary heart disease – a stability substudy, Journal of the American Heart Association ...

    2022 doi

  63. [71]

    G. S. Wei, S. A. Coady, D. C. Go ff Jr, F. L. Brancati, D. Levy, E. Selvin, R. S. Vasan, C. S. Fox, Blood pressure and the risk of developing diabetes in african americans and whites: Aric, cardia, and the framingham heart study, Diabetes care 34 (4) (2011) 873–879. doi:10.233...

    2011 doi

  64. [72]

    E.-J. Rhee, K. Han, S.-H. Ko, K.-S. Ko, W.-Y . Lee, Increased risk for diabetes development in subjects with large variation in total cholesterol levels in 2,827,950 koreans: A nationwide population-based study, PLOS ONE 12 (2017) e0176615. doi:10.1371/journal.pone.0176615. UR...

    2017 doi

  65. [73]

    E. S. of Cardiology, Body mass index is a more powerful risk factor for diabetes than genetics, ScienceDaily (2020). URL www.sciencedaily.com/releases/2020/08/ 200831090129.htm

    2020

  66. [74]

    DiPietro, D

    L. DiPietro, D. Buchner, D. Marquez, R. Pate, L. Pescatello, M. Whitt- Glover, New scientific basis for the 2018 u.s. physical activity guidelines, Journal of Sport and Health Science 8 (2019) 197–200. doi:10.1016/j.jshs.2019.03.007. URL https://www.ncbi.nlm.nih.gov/pmc/articl...

    2019 doi

  67. [75]

    K. Sil, B. K. Das, S. Pal, L. Mandal, A study on impact of education on diabetic control and complications, National Journal of Medical Research 10 (01) (2020) 26–29. URL https://njmr.in/index.php/file/article/view/48

    2020

  68. [76]

    Diabetes, UK, Cardiovascular Disease and Di- abetes, https://www.diabetes.org.uk/ guide-to-diabetes/complications/cardiovascular disease, [Online] Accessed April 15, 2024

    2024

  69. [77]

    Verma, V

    S. Verma, V . Boonsanong, M. Hoang, K. E. Hines, J. P. Dick- erson, C. Shah, Counterfactual explanations and algorithmic re- courses for machine learning: A review (2022). arXiv:2010.10596, doi:https://doi.org/10.48550/arXiv.2010.10596

    2022 doi arXiv

  70. [78]

    S. B. Mushlin, H. L. Greene, Decision making in medicine: an algo- rithmic approach, Elsevier Health Sciences, 2009. doi:10.1016 /S0377- 1237(02)80153-8. 22

    2009

  71. [2399]

    URL https://www.ahajournals.org/doi/full/ 10.1161/circulationaha.106.682658

    doi:https://doi.org/10.1161/CIRCULATIONAHA.106.682658. URL https://www.ahajournals.org/doi/full/ 10.1161/circulationaha.106.682658

    doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.