Pith. sign in

REVIEW 4 major objections 4 minor 62 references

Patient Similarity Computation for Clinical Decision Support: An Efficient Use of Data Transformation, Combining Static and Time Series Data

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Similarity-based heart-disease prediction improves when static data are transformed before clustering.

desk verdict The reported CAD/CHF gains are likely label leakage through aWOE; the Spark DTW engineering is real but the central result is not credible as presented. read the letter →

arxiv 2506.07092 v1 pith:VIMVNMKD submitted 2025-06-08 cs.LG cs.AI

Joydeb Kumar Sana , Mohammad M. Masud , M Sohel Rahman , M Saifur Rahman This is my paper · ORCID
classification cs.LGcs.AI
keywords patientsimilarityclinicaldecisionsupportdynamictimewarpingadaptiveweight-of-evidencedatatransformationdistributedcomputingcardiovasculardiseaseprediction
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

The paper proposes a distributed patient-similarity pipeline that predicts coronary artery disease and congestive heart failure by combining two information sources: static demographic and history features, used to cluster patients, and 18 physiological time series, compared within clusters using Dynamic Time Warping. The paper's central move is to apply data transformations—adaptive Weight-of-Evidence (aWOE) and Z-score—to the static features before clustering. The authors report that this transformation step raises AUC, accuracy, and F-measure by up to 11.4%, 10.20%, and 12.6% for coronary artery disease and up to 15.9%, 10.5%, and 21.9% for congestive heart failure relative to the same pipeline without transformation, while a distributed execution environment cuts computation time by up to 40%. A sympathetic reader should care because this suggests a cheap preprocessing step can make similarity-based diagnosis markedly more accurate and fast enough for real-time clinical decision support.

What carries the argument

The load-bearing object is the neighborhood similarity fusion: for each time-series variate, a DTW distance matrix is computed between a target patient and patients in the same static-data cluster, the per-variate nearest neighbors are unioned into a similarity fusion, and the majority label among those neighbors is assigned. The other named component is adaptive Weight-of-Evidence (aWOE), a binning transform that replaces each static feature value with the log-ratio of positive to negative outcome counts within its bin, plus a small constant; the paper argues this transformation both improves cluster quality and obscures raw patient values, and it reports that aWOE consistently beats Z-score and no-transformation variants.

What would settle it

Re-run the aWOE-K-means pipeline twice, once with aWOE bins computed from the true disease label and once with bins computed from a shuffled binary label assigned randomly to the same patients; if the AUC and F-measure gains shrink or disappear in the shuffled case, the transformation's reported benefit is attributable to target leakage rather than to the transformation itself.

Watch

Extended reading notes

Core claim

The central claim is that the gain in patient-similarity prediction comes from the data transformation, not from the clustering or the time-series distance alone. With the aWOE-K-means variant, the model reaches an AUC of 0.858, accuracy of 0.870, and F-measure of 0.896 for coronary artery disease, and an AUC of 0.878, accuracy of 0.887, and F-measure of 0.816 for congestive heart failure; without any transformation the same pipeline drops to AUC values near 0.61 for both diseases. The paper further reports that a 12-hour observation window recovers almost all of the performance of the full-length time series (F-measure differences of 3.13 and 4.12 for the two diseases), and that adding computing nodes reduces the DTW computation time, making the method viable for real-time use.

Load-bearing premise

The load-bearing premise is that the binary outcome used to compute the aWOE transformation and the disease labels used as clustering features are not the same as, or strongly correlated with, the disease being predicted; if they are, the reported improvements could come from target information leaking into the features rather than from better similarity learning.

Editorial extensions

If this is right

  • If the transformation result holds, a simple preprocessing change—aWOE on static features—can lift the accuracy of similarity-based CAD and CHF prediction above all tested baselines, including LSTM and Transformer models.
  • The 12-hour result implies that intensive-care monitoring windows of half a day are enough for near-optimal patient matching, which matters for emergency settings where full histories are unavailable.
  • The reported runtime scaling, with the full pipeline taking roughly 53 seconds per target patient on a 100-patient cohort, implies distributed DTW can meet real-time decision-support demands as node counts grow.
  • Because the neighborhood fusion ignores a variate when a patient's record is missing it, the method has a built-in tolerance for missing time-series data.

Reading between the lines

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

  • Editorial inference: the aWOE scores in Eq. (1) are computed from distributions of positive and negative events, and the paper does not state that this outcome label is different from the disease being predicted; if the same label feeds the transformation and the clustering features (CHF appears as a clustering feature for CAD and vice versa), part of the reported gain could reflect target leakage
  • Editorial inference: the privacy claim rests on the aWOE binning being a generalization-based anonymity mechanism; a concrete re-identification attack or a k-anonymity audit on the transformed static matrix would test that guarantee directly.
  • Editorial inference: the 40% runtime reduction is measured on one cluster configuration; the released code would allow a reader to check whether the speedup transfers to longer time series or larger cohorts.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes a distributed patient similarity computation (DPSC) method that combines dynamic time warping (DTW) on 18 time series variables with clustering on static patient data. Before clustering, the static data are transformed with either adaptive Weight-of-Evidence (aWOE) or Z-score. For a target patient, the nearest neighbor is selected from the same cluster using a fusion of per-variate DTW distance matrices, and the neighbor's label is used for prediction. The method is evaluated on MIMIC-III for predicting Coronary Artery Disease (CAD) and Congestive Heart Failure (CHF), with claims of large improvements over non-transformed baselines and a 40% runtime reduction from Spark-based distributed DTW. The central claim is that the aWOE transformation and the clustering step improve patient similarity prediction.

Significance. If the reported results are valid, the paper would offer a practical distributed approach to patient similarity that integrates static and time series data with a privacy-oriented transformation, and it would provide a useful benchmark for CAD and CHF prediction on MIMIC-III. The use of a public dataset, the comparison against several baselines, and the availability of code are strengths. However, the significance hinges entirely on whether the aWOE transformation and clustering are free of target leakage; the current manuscript does not establish this, so the headline improvements cannot be taken at face value.

major comments (4)
  1. [Section 2.2.1, Eq. (1)] Equation (1) defines aWOE using the 'Distribution of positive events' and 'Distribution of negative events' but does not specify which binary outcome defines these events. Since Section 2.1.1 lists Coronary Artery Disease and Congestive Heart Failure among the static features, and Section 2.3 uses these disease labels as clustering features, the aWOE transformation may be target-encoding the very outcome being predicted. If the outcome in Eq. (1) is the same as the prediction target, then the clusters from which nearest neighbors are drawn are constructed with label information, and the performance gaps in Table 3 (e.g., aWOE K-means CAD AUC 0.858 vs. No-DT K-means 0.605) would be artifacts of leakage rather than evidence of genuine similarity learning. The manuscript must specify, for each prediction task, the outcome used to compute aWOE and confirm that this outcome is not the target or a direct proxy for it.
  2. [Section 2.3] The text states that for CHF prediction, the Coronary Artery Disease attribute is included as a clustering feature, and for CAD prediction, Congestive Heart Failure is included, with only the target attribute itself disregarded. Because CAD and CHF are strongly correlated cardiovascular conditions, using the other disease label as a clustering feature is a probable channel of label leakage: the cluster assignment, and hence the available pool of nearest neighbors, is influenced by a label that is highly predictive of the target. The authors should report the correlation between the two labels and evaluate a variant that excludes both disease labels from the static feature set used for clustering.
  3. [Section 3.2 and Section 2.9] Section 3.2 describes selecting the number of clusters K based on the average performance over five runs and reports optimal values (K=125 for CAD, K=150 for CHF), but it does not clarify whether this selection used the test split. If the test set was used to choose K, the reported peak performance is optimistically biased. Moreover, Section 2.9 describes an 80/20 train/test split but does not state whether the aWOE binning and the clustering are fitted on the training fold only; if the transformation or clustering uses the full dataset, including test patients, then test labels and feature distributions leak into neighbor selection. Please clarify the validation protocol and, if needed, re-run the evaluation with the selection and fitting steps strictly inside the training folds.
  4. [Section 4, Table 5] The Holm post-hoc table lists the comparison 'aWOE based DPSC vs. Z-score based DPSC' with P-value 0.133614 and marks the hypothesis as 'Rejected', which is inconsistent because 0.133614 exceeds the adjusted alpha of 0.05. In addition, the significance level is given as α=0.5 in the narrative text of Section 4 while Section 2.12 states α=0.05. These inconsistencies undermine the statistical-significance claim and should be corrected.
minor comments (4)
  1. [Section 2.7] The symbol K is used both for the number of clusters and for the number of univariate time series (K=18), which is confusing; please use distinct symbols (e.g., C for clusters and V for variates).
  2. [Table 3] For CAD with No-DT OPTICS, specificity is reported as 0.81 but recall as 0.375, yielding an F-measure of 0.503; a brief explanation of this imbalance would help the reader interpret the row.
  3. [Abstract and Section 3.4] The abstract claims computation time reduction 'by as high as 40%', but Figures 11–13 compare runtime only for DTW distance computation, not for the full DPSC pipeline; please specify the scope of the 40% claim.
  4. [Section 2.7, Eq. (5)] The fusion function F in Eq. (5) is not defined formally; please state how the K distance matrices are combined (e.g., averaging, max, or weighted sum).

Circularity Check

1 steps flagged · score 8.0 of 10

aWOE features are target-encoded by Eq. (1), so the clusters that determine nearest neighbors encode the very label being predicted; the reported gains reduce in substantial part to self-prediction.

  1. self definitional [Section 2.2.1, Eq. (1); applied in Sections 2.3 and 2.9]
    "The mathematical equation of the aWOE can be expressed as: aWOE = ln(Distribution of positive events in a particular bin + ε / Distribution of negative events in a particular bin + ε) (1)"

    Eq. (1) defines each transformed static value as the log-odds of 'positive events' within its bin, and the paper never specifies that 'positive events' is anything other than the CAD/CHF label being predicted. Section 2.3 then clusters patients on these transformed static features and restricts every target patient's neighbors to its own cluster, and Section 2.9 outputs the majority neighbor label as the prediction. Thus cluster membership, hence the selected nearest neighbor and its label, is a function of the label used to compute aWOE. The statement that the CAD attribute is 'disregarded' when predicting CAD removes only the raw label column; the aWOE-transformed versions of age, weight, and other static fields still carry label information.

full rationale

The central derivation chain is: static features → aWOE/Z-score → clustering → cluster-restricted DTW neighbors → majority-label prediction. The aWOE step is the circular pivot: Eq. (1) defines transformed values from the distribution of 'positive events' and 'negative events'; with no other binary outcome defined anywhere in the paper, this is naturally read as the CAD/CHF label the system is supposed to predict. Clustering on these transformed features determines which patients can be neighbors, and the neighbor's label is the output, so the prediction loop closes through the label itself. The paper's claim that it 'disregards' the target attribute in clustering is insufficient, because the aWOE encoding of every static feature already embeds the target distribution. Other contributions are not circular: DTW similarity on time series is computed without target information, Z-score is unsupervised, and the Spark speedup is an engineering result. The self-citation to [43] for aWOE's privacy property is ancillary to the prediction claim and does not change the score. Because the strongest reported gains are tied to the aWOE-transformed static clustering, the headline improvement reduces in substantial part to the fitted label encoding; I assign 8 rather than 10 because Eq. (1)'s outcome variable is not explicitly named and the exact degree of leakage therefore depends on that unspecified choice.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central methodological burden is that the feature construction (aWOE) requires a binary target to compute, and the clustering step uses a correlated disease label, so the method relies on the assumption that these do not leak the prediction label. There are several hand-set free parameters (K, lambda, epsilon, q, window length), and the privacy guarantee is assumed from self-cited work. No new theoretical entities are introduced.

free parameters (6)
  • Number of clusters K = 125 (CAD), 150 (CHF)
    Chosen by maximizing F-measure on test data (Figures 8-9), a form of test-set tuning.
  • aWOE binning parameter q = unspecified
    The number of bins is sample_size / q when unique values exceed 100, but q is never defined in Section 2.2.1.
  • aWOE adjustment constant epsilon = 0.0001
    Hand-set constant added to avoid log(0) in Eq. (1).
  • aWOE unique-value threshold lambda = 100
    Hand-set threshold for deciding binning strategy in Section 2.2.1.
  • Number of nearest neighbors lambda = 1
    Nearest neighbor count used in Section 2.8; no sensitivity analysis for lambda.
  • Observation window length = 12 hours
    Selected because it gave near-optimal F-measure while reducing computation (Section 3.3).
assumptions (4)
  • domain assumption aWOE transformation can be computed without using the prediction target, so no label leakage occurs
    Eq. (1) uses distributions of positive and negative events, which are defined by a binary outcome; the paper does not state this outcome is excluded from feature construction, and Section 2.3 uses a related disease label in clustering.
  • domain assumption Clustering on all patients before the 80/20 train/test split does not leak test information
    Section 2.9 describes the split after clustering, so test patients participate in cluster formation; this is transductive and can inflate performance if cluster structure is label-related.
  • domain assumption DTW distance is an appropriate similarity measure for multivariate, unevenly sampled vital-sign time series
    DTW assumes meaningful alignment; unequal sampling and missing values are handled by ignoring empty neighborhoods, but this is not evaluated.
  • domain assumption aWOE preserves privacy equivalently to k-anonymity
    Section 7 relies on prior work [43] by the same authors; no formal privacy proof or empirical privacy metric is given in this paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Patient Similarity Computation for Clinical Decision Support: An Efficient Use of Data Transformation, Combining Static and Time Series Data." pith.science (2026). https://pith.science/paper/VIMVNMKD

@misc{pith2026250607092,
  author       = {Pith},
  title        = {Pith review of: Patient Similarity Computation for Clinical Decision Support: An Efficient Use of Data Transformation, Combining Static and Time Series Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VIMVNMKD}},
  note         = {Machine review of arXiv:2506.07092}
}
read the original abstract

Patient similarity computation (PSC) is a fundamental problem in healthcare informatics. The aim of the patient similarity computation is to measure the similarity among patients according to their historical clinical records, which helps to improve clinical decision support. This paper presents a novel distributed patient similarity computation (DPSC) technique based on data transformation (DT) methods, utilizing an effective combination of time series and static data. Time series data are sensor-collected patients' information, including metrics like heart rate, blood pressure, Oxygen saturation, respiration, etc. The static data are mainly patient background and demographic data, including age, weight, height, gender, etc. Static data has been used for clustering the patients. Before feeding the static data to the machine learning model adaptive Weight-of-Evidence (aWOE) and Z-score data transformation (DT) methods have been performed, which improve the prediction performances. In aWOE-based patient similarity models, sensitive patient information has been processed using aWOE which preserves the data privacy of the trained models. We used the Dynamic Time Warping (DTW) approach, which is robust and very popular, for time series similarity. However, DTW is not suitable for big data due to the significant computational run-time. To overcome this problem, distributed DTW computation is used in this study. For Coronary Artery Disease, our DT based approach boosts prediction performance by as much as 11.4%, 10.20%, and 12.6% in terms of AUC, accuracy, and F-measure, respectively. In the case of Congestive Heart Failure (CHF), our proposed method achieves performance enhancement up to 15.9%, 10.5%, and 21.9% for the same measures, respectively. The proposed method reduces the computation time by as high as 40%.

Figures

Figures reproduced from arXiv: 2506.07092 by the authors.

Figure 1
Figure 1. High level block diagram of DT based distributed patient similarity model (DT-DPSM) [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. The box plot diagram shows the distribution of the number of data points in various time series data items, excluding [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. T and S represent two time series for a specific variable v across the time axis t. The Euclidean distance is calculated [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (12 more)
Figure 4
Figure 4. Figure 4: Time series distance measurement using DTW. [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Warping path of DTW [42] supporting general batch processing, streaming analytics, machine learning, graph analytics, ad-hoc queries, and more. It has become one of the most popular tools for running analytics jobs and mainstream solution for big data analytics [46][47…
Figure 6
Figure 6. Figure 6: illustrates the overall computation flow for the distributed patient similarity model. Firstly, Using the clustering algorithms, patients are partitioned into groups. Before using the static data into the clustering algorithms, DT techniques have been applied on the da…
Figure 7
Figure 7. Figure 7: Comparison between DT-clustering based, No-DT-clustering based and Non-clustering based models. [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]
Figure 8
Figure 8. Figure 8: Cluster count versus prediction performance using [PITH_FULL_IMAGE:figures/full_fig_p016_8.png]
Figure 9
Figure 9. Figure 9: Cluster count versus prediction performance using [PITH_FULL_IMAGE:figures/full_fig_p016_9.png]
Figure 10
Figure 10. Figure 10: Prediction performance utilizing various time series data lengths in terms of F-measure. [PITH_FULL_IMAGE:figures/full_fig_p017_10.png]
Figure 11
Figure 11. Figure 11: Comparison: Execution time Vs. Time-series data length [PITH_FULL_IMAGE:figures/full_fig_p017_11.png]
Figure 12
Figure 12. Figure 12: Comparison: Execution time Vs. number of patient size [PITH_FULL_IMAGE:figures/full_fig_p018_12.png]
Figure 13
Figure 13. Figure 13: Comparison: Execution time Vs. number of Spark nodes [PITH_FULL_IMAGE:figures/full_fig_p018_13.png]
Figure 14
Figure 14. Figure 14: The Beeswarm SHAP plot using K-means Clustering in terms of Coronary Artery Disease. [43], aWOE serves as a privacy preserving mechanism similar to k-anonymity. In particular, it employs a generalization-based k-anonymity approach in its transformation process. During…
Figure 15
Figure 15. Figure 15: The Beeswarm SHAP plot using K-means Clustering in terms of Congestive Heart Failure. The previously mentioned work [34] studied the effectiveness of the combination of the time series data and static data. However, they did not consider the data transformation method…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

62 extracted references · 55 canonical work pages

  1. [1]

    Patient similarity in prediction mod- els based on health data: A scoping review

    Anis Sharafoddini, Joel A Dubin, J.L., 2017. Patient similarity in prediction mod- els based on health data: A scoping review. JMIR medical informatics 5. URL: https://pubmed.ncbi.nlm.nih.gov/28258046/, doi:https://doi.org/10.2196/medinform.6730

    work page arXiv 2017

  2. [3]

    Ankerst, M., Breunig, M.M., Kriegel, H.P., Sander, J., 1999b. Optics: ordering points to identify the clustering structure, in: Proceedings of the 1999 ACM SIGMOD International Conference on Management of Data, Association for Computing Machinery, New York, NY, USA. p. 49–60. URL: https://doi.org/10.1145/304182.304187, doi:10.1145/304182.304187

    arXiv 1999

  3. [4]

    Using dynamic time warping to find patterns in time series, in: In Pro- ceedings of the 3rd International Conference on Knowledge Discovery and Data Mining (AAAIWS’94), p

    Berndt, D.J., Clifford, J., 1994. Using dynamic time warping to find patterns in time series, in: In Pro- ceedings of the 3rd International Conference on Knowledge Discovery and Data Mining (AAAIWS’94), p. 359–370. doi:https://dl.acm.org/doi/10.5555/3000850.3000887

    work page arXiv 1994

  4. [5]

    Real-time prediction of mortality, readmission, and length of stay us- ing electronic health record data

    Cai, X., Perez-Concha, O., Coiera, E., Martin-Sanchez, F., Day, R., Roffe, D., Gal- lego, B., 2015. Real-time prediction of mortality, readmission, and length of stay us- ing electronic health record data. Journal of the American Medical Informatics Associa- tion 23, 553–561. URL:https://doi.org/10.1093/jamia/ocv110, doi:10.1093/jamia/ocv110, arXiv:https:...

    work page doi:10.1093/jamia/ocv110 2015

  5. [6]

    Similarity Measures and Dimensionality Reduction Techniques for Time Series Data Mining

    Cassisi, C., Montalto, P., Aliotta, M., Cannata, A., Pulvirenti, A., 2012. Similarity Measures and Dimensionality Reduction Techniques for Time Series Data Mining. doi:10.5772/49941

    work page doi:10.5772/49941 2012

  6. [7]

    Chan, L., Chan, T., Cheng, L., Mak, W., 2010. Machine learning of patient similarity: A case study on predicting survival in cancer patient after locoregional chemotherapy, in: 2010 IEEE International Conference on Bioinformatics and Biomedicine Workshops (BIBMW), pp. 467–470. doi:10.1109/BIBMW.2010.5703846

    work page arXiv 2010

  7. [8]

    Exploiting Convolutional Neural Network for Risk Prediction with Medical Feature Embedding

    Che, Z., Cheng, Y., Sun, Z., Liu, Y., 2017. Exploiting convolutional neural network for risk prediction with medical feature embedding. ArXiv abs/1701.07474. 22

    work page Pith review arXiv 2017

  8. [9]

    Analysis of microarray data using z-score trans- formation

    Cheadle, C., Vawter, M., Freed, W., Becker, K., 2003. Analysis of microarray data using z-score trans- formation. The Journal of molecular diagnostics : JMD 5, 73–81. doi:10.1016/S1525-1578(10)60455-2

    work page doi:10.1016/s1525-1578(10)60455-2 2003

Show all 62 references
  1. [10]

    Risk Prediction with Elec- tronic Health Records: A Deep Learning Approach

    Cheng, Y., Wang, F., Zhang, P., Hu, J., . Risk Prediction with Elec- tronic Health Records: A Deep Learning Approach. pp. 432–440. URL: https://epubs.siam.org/doi/abs/10.1137/1.9781611974348.49, doi:10.1137/1.9781611974348.49, arXiv:https://epubs.siam.org/doi/pdf/10.1137/1.978...

    doi

  2. [11]

    Introduction to Algorithms, Third Edition

    Cormen, T.H., Leiserson, C.E., Rivest, R.L., Stein, C., 2009. Introduction to Algorithms, Third Edition. 3rd ed., The MIT Press

    2009

  3. [12]

    A comparative analy- sis of data preparation algorithms for customer churn prediction: A case study in the telecommunication industry

    Coussement, K., Lessmann, S., Verstraeten, G., 2017. A comparative analy- sis of data preparation algorithms for customer churn prediction: A case study in the telecommunication industry. Decision Support Systems 95, 27 – 36. URL:http://www.sciencedirect.com/science/article/pi...

    2017 doi

  4. [13]

    Generating evidence based interpretation of hematology screens via anomaly charac- terization

    David, G., 2011. Generating evidence based interpretation of hematology screens via anomaly charac- terization. The Open Clinical Chemistry Journal 4, 10–16. doi:10.2174/1874241601104010010

    2011 doi

  5. [14]

    Statistical comparisons of classifiers over multiple data sets

    Demˇ sar, J., 2006. Statistical comparisons of classifiers over multiple data sets. Journal of Machine Learning Research 7, 1–30

    2006

  6. [15]

    Exact indexing of dynamic time warping

    Eamonn Keogh, C.A.R., 2002. Exact indexing of dynamic time warping. In proceedings of the 26th Int’l Conference on Very Large Data Bases. Hong Kong , 406–417

    2002

  7. [16]

    Tsiklidis, Talid Sinno, S.L.D., 2022

    Evan J. Tsiklidis, Talid Sinno, S.L.D., 2022. Predicting risk for trauma patients us- ing static and dynamic information from the mimic iii database. PLoS ONE 17. URL:https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0262523, doi:https://doi.org/10.1371/journal...

    2022 doi

  8. [17]

    Fast subsequence matching in time-series databases

    Faloutsos, C., Ranganathan, M., Manolopoulos, Y., 1994. Fast subsequence matching in time-series databases. ACM SIGMOD Record 23, 419 – 429. doi:https://doi.org/10.1145/191843.191925

    1994

  9. [18]

    Intertemporal similarity of economic time series: An application of dynamic time warping

    Franses, P.H., Wiemann, T., 2020. Intertemporal similarity of economic time series: An application of dynamic time warping. Computational Economics 56, 59–75. doi:https://doi.org/10.1007/s10614-020- 09986-0

    2020 doi

  10. [19]

    Patient clustering with uncoded text in electronic medical records

    Henao, R., Murray, J., Ginsburg, G., Carin, L., Lucas, J.E., 2013. Patient clustering with uncoded text in electronic medical records. AMIA ... Annual Symposium proceedings. AMIA Symposium 2013, 592–9

    2013

  11. [20]

    Survey of clinical data mining applications on big data in health informatics, in: 2013 12th International Conference on Machine Learning and Applica- tions, pp

    Herland, M., Khoshgoftaar, T.M., Wald, R., 2013. Survey of clinical data mining applications on big data in health informatics, in: 2013 12th International Conference on Machine Learning and Applica- tions, pp. 465–472. doi:10.1109/ICMLA.2013.163

    2013 doi

  12. [21]

    Using participant similarity for the classification of epidemiological data on hepatic steatosis, IEEE

    Hielscher, T., Spiliopoulou, M., Volzke, H., Kuhn, J.P., 2014. Using participant similarity for the classification of epidemiological data on hepatic steatosis, IEEE. pp. 1–7. doi:10.1109/CBMS.2014.28

    2014 doi

  13. [22]

    Spectral clustering strategies for heterogeneous disease expression data

    Huang, G.T., Cunningham, K.I., Benos, P.V., Chennubhotla, C.S., 2013. Spectral clustering strategies for heterogeneous disease expression data. Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing , 212–23

    2013

  14. [23]

    Dtw-nn: A novel neural network for time se- ries recognition using dynamic alignment between inputs and weights

    Iwana, B.K., Frinken, V., Uchida, S., 2020. Dtw-nn: A novel neural network for time se- ries recognition using dynamic alignment between inputs and weights. Knowledge-Based Systems 188, 104971. URL:https://www.sciencedirect.com/science/article/pii/S0950705119303995, doi:https:...

    2020

  15. [24]

    A patient-similarity-based model for diagnostic prediction

    Jia, Z., Zeng, X., Duan, H., Lu, X., Li, H., 2020. A patient-similarity-based model for diagnostic prediction. International Journal of Medical Informatics 135, 104073. URL:https://www.sciencedirect.com/science/article/pii/S1386505619310925, doi:https://doi.org/10.1016/j.ijmed...

    2020

  16. [25]

    A novel customer churn prediction model for the telecommunication industry us- ing data transformation methods and feature selection

    Joydeb Kumar Sana, Mohammad Zoynul Abedin, M.S.R., Rahman, M.S., 2022. A novel customer churn prediction model for the telecommunication industry us- ing data transformation methods and feature selection. PLoS ONE 17. URL: https://journals.plos.org/plosone/article?id=10.1371/j...

    2022 doi

  17. [26]

    An efficient k- means clustering algorithm: analysis and implementation

    Kanungo, T., Mount, D., Netanyahu, N., Piatko, C., Silverman, R., Wu, A., 2002. An efficient k- means clustering algorithm: analysis and implementation. IEEE Transactions on Pattern Analysis and Machine Intelligence 24, 881–892. doi:10.1109/TPAMI.2002.1017616

    2002 arXiv

  18. [27]

    Integrated optimisation method for personalised modelling and case studies for medical decision support

    Kasabov, N.K., Hu, Y., 2010. Integrated optimisation method for personalised modelling and case studies for medical decision support. Int. J. Funct. Informatics Pers. Medicine 3, 236–256

    2010

  19. [28]

    Exact indexing of dynamic time warping, in: Proceedings of the 28th International Conference on Very Large Data Bases, VLDB Endowment

    Keogh, E., 2002. Exact indexing of dynamic time warping, in: Proceedings of the 28th International Conference on Very Large Data Bases, VLDB Endowment. p. 406–417

    2002

  20. [29]

    A fast and accurate similarity measure for long time series classification based on local extrema and dynamic time warping

    Lahreche, A., Boucheham, B., 2021. A fast and accurate similarity measure for long time series classification based on local extrema and dynamic time warping. Expert Systems with Applications 168, 114374. URL:https://www.sciencedirect.com/science/article/pii/S0957417420310514,...

    2021

  21. [30]

    Medical time series classification with hierarchical attention-based temporal convolutional networks: A case study of myotonic dystrophy diagnosis

    Lin, L., Xu, B., Wu, W., Richardson, T.W., Bernal, E.A., 2019. Medical time series classification with hierarchical attention-based temporal convolutional networks: A case study of myotonic dystrophy diagnosis. ArXiv abs/1903.11748. URL:https://api.semanticscholar.org/CorpusID...

    2019 arXiv

  22. [31]

    A ¡i¿k¡/i¿ -nearest neighbors survival probability prediction method

    Lowsky, D., Ding, Y., Lee, D., McCulloch, C., Ross, L., Thistlethwaite, J., Zenios, S., 2013. A ¡i¿k¡/i¿ -nearest neighbors survival probability prediction method. Statistics in Medicine 32, 2062–

    2013

  23. [32]

    Integrate multi-omic data using affinity network fusion (anf) for cancer patient clustering, in: 2017 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp

    Ma, T., Zhang, A., 2017. Integrate multi-omic data using affinity network fusion (anf) for cancer patient clustering, in: 2017 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 398–403. doi:10.1109/BIBM.2017.8217682

    2017

  24. [33]

    An integrated data mining approach to real-time clinical monitoring and deterioration warning, pp

    Mao, Y., Chen, W., Chen, Y., Lu, C., Kollef, M., Bailey, T.C., . An integrated data mining approach to real-time clinical monitoring and deterioration warning, pp. 1140–1148. doi:10.1145/2339530.2339709

  25. [34]

    Effective patient similarity computa- tion for clinical decision support using time series and static data, pp

    Masud, M., Hayawi, K., Mathew, S., Dirir, A., Cheratta, M., 2020. Effective patient similarity computa- tion for clinical decision support using time series and static data, pp. 1–8. doi:10.1145/3373017.3373050

    2020

  26. [35]

    Personalized predictive modeling and risk factor identification using patient similarity

    Ng, K., Sun, J., Hu, J., Wang, F., 2015. Personalized predictive modeling and risk factor identification using patient similarity. AMIA Summits on Translational Science Proceedings 2015, 132 – 136

    2015

  27. [36]

    Patient similarity networks for preci- sion medicine

    Pai, S., Bader, G.D., 2018. Patient similarity networks for preci- sion medicine. Journal of Molecular Biology 430, 2924–2938. URL: https://www.sciencedirect.com/science/article/pii/S0022283618305321, doi:https://doi.org/10.1016/j.jmb.2018.05.037. theory and Application of Net...

    2018 doi

  28. [37]

    Patient similarity for precision medicine: A systematic review

    Parimbelli, E., Marini, S., Sacchi, L., Bellazzi, R., 2018. Patient similarity for precision medicine: A systematic review. Journal of Biomedical Informatics 83. doi:10.1016/j.jbi.2018.06.001

    2018 doi

  29. [38]

    New knowledge extraction technique using prob- ability for case-based reasoning: application to medical diagnosis

    Park, Y.J., Kim, B.C., Chun, S.H., 2006. New knowledge extraction technique using prob- ability for case-based reasoning: application to medical diagnosis. Expert Systems 23, 2–

    2006

  30. [39]

    Physionet-MIMICIII

    Physionet, . Physionet-MIMICIII. [n.d.].https://archive.physionet.org/physiobank/database/mimic3cdb/. Accessed: November 22, 2021

    2021

  31. [40]

    URL:https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1468-0394.2006.00321.x, doi:https://doi.org/10.1111/j.1468-0394.2006.00321.x,arXiv:https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1468-0394.2006.00321.x

    2006

  32. [41]

    netdx: interpretable patient classification using integrated patient similarity networks

    S, P., S, H., R, I., MA, S., H, K., GD., B., 2019. netdx: interpretable patient classification using integrated patient similarity networks. Mol Syst Biol. 15, 406–417. doi:doi:10.15252/msb.20188497. 24

    2019 doi

  33. [42]

    Benchmarking deep learn- ing models on large healthcare datasets

    Purushotham, S., Meng, C., Che, Z., Liu, Y., 2018. Benchmarking deep learn- ing models on large healthcare datasets. Journal of Biomedical Informatics 83, 112–

    2018

  34. [43]

    Privacy-preserving customer churn prediction model in the context of telecommunication industry

    Sana, J.K., Rahman, M.S., Rahman, M.S., 2024. Privacy-preserving customer churn prediction model in the context of telecommunication industry. URL:https://arxiv.org/abs/2411.01447, arXiv:2411.01447

    2024 arXiv

  35. [44]

    Modeling online customer purchase intention behavior applying dif- ferent feature engineering and classification techniques

    Satu, M.S., Islam, S.F., 2023. Modeling online customer purchase intention behavior applying dif- ferent feature engineering and classification techniques. Discover Artificial Intelligence 3. URL: https://doi.org/10.1007/s44163-023-00086-0, doi:10.1007/s44163-023-00086-0

    2023 doi

  36. [45]

    Toward accurate dynamic time warping in linear time and space

    Salvador, S., Chan, P., 2007. Toward accurate dynamic time warping in linear time and space. Intell. Data Anal. 11, 561–580

    2007

  37. [46]

    Spark.https://spark.apache.org/

    Spark, a. Spark.https://spark.apache.org/. Accessed on 15.11.2021

    2021

  38. [47]

    Spark.https://www.infoworld.com/article/3236869/what-is-apache-spark-the-big-data-platform-that-crushed-hadoop.html

    Spark, b. Spark.https://www.infoworld.com/article/3236869/what-is-apache-spark-the-big-data-platform-that-crushed-hadoop.html. Accessed on 15.11.2021

    2021

  39. [48]

    Clustering patients ac- cording to health perceptions: Relationships to psychosocial characteristics and medication nonadherence

    Sewitch, M.J., Leffondr´ e, K., Dobkin, P.L., 2004. Clustering patients ac- cording to health perceptions: Relationships to psychosocial characteristics and medication nonadherence. Journal of Psychosomatic Research 56, 323–332. URL:https://www.sciencedirect.com/science/articl...

    2004 doi

  40. [49]

    Deep patient similarity learning for personalized healthcare

    Suo, Q., Ma, F., Yuan, Y., Huai, M., Zhong, W., Gao, J., Zhang, A., 2018. Deep patient similarity learning for personalized healthcare. IEEE Transactions on NanoBioscience PP, 1–1. doi:10.1109/TNB.2018.2837622

    2018

  41. [50]

    Attention is all you need, in: Proceedings of the 31st International Conference on Neural Information Processing Systems, Curran Associates Inc., Red Hook, NY, USA

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, L., Polosukhin, I., 2017. Attention is all you need, in: Proceedings of the 31st International Conference on Neural Information Processing Systems, Curran Associates Inc., Red Hook, NY, USA. p...

    2017

  42. [51]

    A system for mining temporal physiological data streams for advanced prognostic decision support, IEEE

    Sun, J., Sow, D., Hu, J., Ebadollahi, S., 2010. A system for mining temporal physiological data streams for advanced prognostic decision support, IEEE. pp. 1061–1066. doi:10.1109/ICDM.2010.102

    2010 doi

  43. [52]

    Wang, B., Mezlini, A., Demir, F., Fiume, M., Tu, Z., Brudno, M., Haibe-Kains, B., Goldenberg, A.,

  44. [53]

    Wang, F., Hu, J., Sun, J., 2012a. Medical prognosis based on patient similarity and expert feedback, in: 2012 21st International Conference on Pattern Recognition (ICPR 2012), IEEE Computer Society, Los Alamitos, CA, USA. pp. 1799–1802. URL:https://doi.ieeecomputersociety.org/

    2012

  45. [54]

    Comparing and combining time series trajectories using dynamic time warping

    Vaughan, N., Gabrys, B., 2016. Comparing and combining time series trajectories using dynamic time warping. Procedia Computer Science 96, 465–474. doi:10.1016/j.procs.2016.08.106

    2016 doi

  46. [55]

    Integrating distance metrics learned from mul- tiple experts and its application in inter-patient similarity assessment, in: SDM

    Wang, F., Sun, J., Ebadollahi, S., 2011. Integrating distance metrics learned from mul- tiple experts and its application in inter-patient similarity assessment, in: SDM. URL: https://api.semanticscholar.org/CorpusID:18546025

    2011

  47. [56]

    Multiclass spectral clustering, in: Proceedings Ninth IEEE International Conference on Computer Vision, pp

    Yu, Shi, 2003. Multiclass spectral clustering, in: Proceedings Ninth IEEE International Conference on Computer Vision, pp. 313–319 vol.1. doi:10.1109/ICCV.2003.1238361

    2003 arXiv

  48. [57]

    Hierarchical Agglomerative Clustering

    Zepeda-Mendoza, M.L., Resendis-Antonio, O., 2013. Hierarchical Agglomerative Clustering. Springer New York, New York, NY

    2013

  49. [58]

    Medical prognosis based on patient similarity and expert feedback, in: Proceedings of the 21st International Conference on Pattern Recognition (ICPR2012), pp

    Wang, F., Hu, J., Sun, J., 2012b. Medical prognosis based on patient similarity and expert feedback, in: Proceedings of the 21st International Conference on Pattern Recognition (ICPR2012), pp. 1799–1802

  50. [60]

    Measuring patient similarities via a deep architecture with medical concept embedding, in: 2016 IEEE 16th International Conference on Data Mining (ICDM), pp

    Zhu, Z., Yin, C., Qian, B., Cheng, Y., Wei, J., Wang, F., 2016b. Measuring patient similarities via a deep architecture with medical concept embedding, in: 2016 IEEE 16th International Conference on Data Mining (ICDM), pp. 749–758. doi:10.1109/ICDM.2016.0086. 25

    2016

  51. [62]

    A novel patient similarity prediction model based on semisupervised learning, in: CAIBDA 2022: 2nd International Conference on Artificial Intelligence, Big Data and Algorithms, pp

    Zhang, L., Li, X., Bi, X., Zhang, Y., Yu, G., Zhao, K., 2022. A novel patient similarity prediction model based on semisupervised learning, in: CAIBDA 2022: 2nd International Conference on Artificial Intelligence, Big Data and Algorithms, pp. 1–9

    2022

  52. [134]

    URL:https://www.sciencedirect.com/science/article/pii/S1532046418300716, doi:https://doi.org/10.1016/j.jbi.2018.04.007

    2018 doi

  53. [2014]

    Nature methods 11

    Similarity network fusion for aggregating data types on a genomic scale. Nature methods 11. doi:10.1038/nmeth.2810

    doi

  54. [2069]

    doi:10.1002/sim.5673

    doi

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.