Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Semi-supervised Clustering Through Representation Learning of Large-scale EHR Data

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read SCORE's error rate for EHR phenotyping stops depending on the number of labeled patients.

desk verdict SCORE is a credible semi-supervised phenotyping method with a genuine diverging-q theoretical result, but the central transfer-learning claim rests on an untested assumption about V spanning the true latent space. read the letter →

arxiv 2505.20731 v1 pith:VAI3IYGF submitted 2025-05-27 stat.ME cs.LG

Linshanshan Wang , Mengyan Li , Zongqi Xia , Molei Liu , Tianxi Cai This is my paper · ORCID
classification stat.MEcs.LG MSC 62H3062H2562F1262P10
keywords factormixturemodelembeddingGaussianvariationalapproximationExpectation-MaximizationalgorithmEHRcodifieddatasemi-supervisedlearningpatientphenotypingrepresentation
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 SCORE, a semi-supervised framework that clusters patients and learns patient embeddings from sparse, high-dimensional EHR count data when labeled examples are scarce. It models the counts with a Poisson-adapted latent factor mixture (PALM) that treats pre-trained EHR code embeddings as a known linear subspace, so only scaling parameters, not the full loading matrix, must be estimated. The central theoretical claim is that a hybrid EM and Gaussian variational approximation algorithm, initialized from the supervised fit, converges at a rate dominated by the total sample size $N$ and the feature dimension $p$, with the labeled sample size $n$ absent from the leading term. If correct, label scarcity stops being the bottleneck: large unlabeled pools substitute for labels, and the learned embeddings transfer to related phenotypes such as EDSS-defined disability. Simulations and a multiple sclerosis EHR application support this conclusion.

What carries the argument

The engine is the Poisson-Adapted Latent factor Mixture (PALM): $X\mid Z\sim\mathrm{Poisson}(e^Z)$ with conditionally independent coordinates, $Z=V(B\tilde U_Y+W)$, $W\sim N(0,\Lambda)$, and $Y\mid U\sim\mathrm{Bernoulli}(\mathrm{expit}(b^T\tilde U))$. The matrix $V\in\mathbb{R}^{p\times q}$ is the fixed pre-trained embedding matrix of the $p$ EHR codes; treating its column space as known cuts the loading parameters from $pq$ to $q^2$ scaling parameters. The algorithm then uses a Gaussian variational approximation (GVA) with diagonal covariance per subject to obtain a closed-form ELBO, and the hybrid EM-GVA alternates between imputing the latent class probabilities $\gamma^{\mathrm{GVA}}$ for unlabeled subjects and maximizing the full-sample ELBO, initialized by the supervised ELBO estimator. The convergence proof runs through a linear-contraction lemma for the EM-GVA iterations, which requires the SNR $\psi^2$ to exceed a constant and ties the final rate to $N$ and $p$ rather than to $n$.

What would settle it

Run the PALM simulation with a known $V$, then rerun SCORE with $V$ altered by swapping one column for a random vector while holding $n$ fixed and doubling $N$; if the estimation error stops falling and stays above the predicted $\psi q^{1/2}/N^{1/2}$ term, the span assumption is what carries the result.

Watch

Extended reading notes

Core claim

On the paper's own terms, the SCORE estimator recovers the PALM parameters $\theta=(B,\Lambda,b)$ at rate $$\mathrm{Err}(\hat{\$\theta$})=O_P(\psi $q^{{1/2}}$/$N^{{1/2}}$+\psi $q^{{2-\varepsilon}}$/$p^{{1-\varepsilon}}$),$$ which contains no labeled-sample-size term $n$, provided Assumptions 1-3 hold and the signal-to-noise ratio satisfies $\psi^2>C_\psi$ (Theorem 2). The estimator is built so that the supervised ELBO fit initializes the hybrid EM-GVA procedure and its influence decays geometrically, leaving only the unlabeled-data and GVA-approximation errors. For a new subject, the phenotyping probability $\gamma_{\mathrm{new}}^{\mathrm{GVA}}$ converges to $\Pr(Y=1\mid \bar\xi_{\mathrm{new}},U_{\mathrm{new}})$ at rate $O_P(\psi^2 q^{1/2}/N^{1/2}+\psi q^{1/2-\varepsilon}/p^{1/2-\varepsilon})$, and the estimated embedding $\hat E_{\mathrm{new}}$ converges to the true latent embedding $\bar\xi_{\mathrm{new}}$ (Corollary 1). Because Proposition 1 shows $\Pr(Y=1\mid\bar\xi,U)$ is at least as informative about $Y$ as the raw-count posterior, the learned low-dimensional representations are claimed to be the right object to use for downstream phenotyping and prediction.

Load-bearing premise

The load-bearing premise is that the pre-trained code embeddings already span the true latent disease structure of the target cohort; if those external embeddings point in the wrong directions, the model is misspecified and the labeled-sample-free error rate no longer follows.

Editorial extensions

If this is right

  • Adding unlabeled patients shrinks parameter estimation error at a $\psi q^{1/2}/N^{1/2}$ rate, so a cohort with only dozens of labeled patients can still be phenotyped when the unlabeled pool is large.
  • Only $O(\log(N/n))$ EM iterations are needed to reach that rate, so the benefit of unlabeled data comes at a modest computational cost.
  • For new subjects, the predicted disability probability converges to $\Pr(Y=1\mid\bar\xi,U)$, which Proposition 1 shows is no less informative than using raw count features.
  • The model treats observed labels as noisy proxies for latent cluster membership, which supports cross-scale phenotyping: embeddings trained on PDDS-defined disability generalize to EDSS-defined disability and predict future progression.
  • Any consistent supervised estimator suffices for initialization, so the final error rate does not depend on which supervised baseline is used to start EM.

Reading between the lines

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

  • Editorial extension: the $n$-free rate implies a scaling law for rare-disease studies - enlarge $N$ by pooling visits or collaborating centers and even very small label sets may be enough; the paper does not turn this into a recruitment rule.
  • Editorial extension: the span assumption can be stress-tested directly by deleting a column of $V$ aligned with the known outcome direction and seeing whether the GVA error floor $\psi q^{2-\varepsilon}/p^{1-\varepsilon}$ remains as $N$ grows.
  • Editorial extension: the same contraction argument should carry to other exponential-family mixtures with known feature embeddings, such as negative binomial or zero-inflated counts, which the paper mentions only as future work.
  • Editorial extension: because unlabeled patients do not need an outcome, one learned embedding could serve several downstream labels simultaneously, so a single SCORE fit might phenotype multiple related conditions from one unlabeled pool.
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

3 major / 4 minor

Summary. The paper proposes SCORE, a semi-supervised representation-learning framework for high-dimensional EHR count data. The model is a Poisson-adapted latent factor mixture (PALM) in which a pre-trained code-embedding matrix V is used as a known loading subspace, and only the coefficient matrix B, the random-effect covariance Λ, and the label model b are estimated. Estimation is carried out by a hybrid EM algorithm that uses a supervised ELBO-based estimator for initialization and Gaussian variational approximations for the unlabeled subjects. The main theoretical results are a linear-contraction lemma for the EM-GVA iterates (Lemma 1), a convergence rate for the semi-supervised estimator that is free of the labeled sample size n (Theorem 2), and a rate for the predicted phenotype probability and patient embedding for a new subject (Corollary 1). The empirical section reports simulation studies and an application to multiple sclerosis disability phenotyping in 16,091 patients, claiming improved AUC, PRAUC, and Brier score over several baselines, especially under label scarcity.

Significance. If the theoretical results hold, the paper would be a useful contribution: it extends the theory of Gaussian variational approximation to a low-rank Poisson mixture with diverging latent dimension and semi-supervised data, and it gives a concrete transfer-learning construction based on pre-trained EHR code embeddings. The explicit SNR condition and the n-free N^{-1/2} rate are interesting and would strengthen the case for using unlabeled EHR data in rare-disease settings. The simulation design with 200 replications and the real MS application are also valuable, and the paper is honest about the limitation that the external embedding structure may not match the target population. However, the manuscript as submitted cannot be fully verified because the proofs of the main theorems are deferred to an appendix that is not included, and the robustness analysis does not address the most threatening deviation from the key assumption that V spans the true latent subspace.

major comments (3)
  1. [Section 3, Theorem 2 and Lemma 1] The main theoretical claims—the linear contraction of the EM-GVA algorithm and the n-free convergence rate of the SCORE estimator—are stated without proofs in the main text, and the appendix containing the proofs is not present in the submitted version. Since these results are the paper's central contribution, the manuscript currently does not allow independent verification of the derivation. The authors should either include the full proofs in an appendix or provide a detailed proof sketch in the main text, with the omitted technical lemmas clearly stated.
  2. [Section 4.1, misspecification setting (e), and Section 6] The misspecification simulations are constructed as Z_i = V( eB eD_i + W_i) with the same pre-trained V, so the additional unobserved factor C_i lies inside the column space of V. This does not test the case where the true latent structure has a component orthogonal to span(V), which is the principal threat to Assumption 2(i). The Discussion concedes that when the external and target factor structures diverge substantially, methods beyond rescaling eigenvalues may be needed. Because the practical claim that pre-trained embeddings make SCORE robust to label scarcity depends on this alignment, the manuscript should add simulations with a rotated, truncated, or contaminated V, or provide a formal analysis of the misalignment error.
  3. [Sections 4.1 and 5] The empirical comparisons are not fully reproducible from the information given. The manuscript does not specify the values or tuning of K_B and K_M, the number of EM iterations T, learning rates, convergence criteria, or the choice of q in the application. For the benchmark methods (Lasso, random forest, XGBoost, NN(sup), NN(VAE)), the tuning procedure, network architecture, and training details are omitted. Since the paper claims superior finite-sample performance, these implementation details should be reported or the code should be made available.
minor comments (4)
  1. [Section 2.2, Eq. (3)] The notation for A_{ij}^{(y)} is introduced as E_{W~φ_i^{(y)}} exp{V_j^T(B U_{iy} + W)}, but the superscript y is dropped in parts of the displayed expression; please make the dependence on y explicit throughout the equation and its surrounding text.
  2. [Section 4.1, Table 1 and text] There are minor typos: 'Chique2018' in Table 1 should be 'Chiquet et al. [2018]', and 'preresentation' should be 'representation' in the sentence about patient representation quality.
  3. [Section 5] The aggregation window of 6 months before and after each visit is a free parameter, but no sensitivity analysis is reported; please justify this choice or show that the results are not sensitive to it.
  4. [Remark 1] The statement that allowing Λ to be freely estimated adjusts for 'scale heterogeneity as well as the rotation ambiguity' is informal; a precise statement of which aspects of the loading matrix are identifiable when V is fixed would help the reader understand the model's identification.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SCORE's derivation is self-contained given the fixed embedding subspace V and stated assumptions; self-citations supply input features rather than load-bearing justification.

full rationale

The paper's derivation chain treats the pre-trained embedding matrix V as a fixed, known input (Section 2.1, model (1), with sqrt(q/p)V orthonormal in Assumption 2(i)); it is not fitted to the target outcome in the theory, and the convergence rates in Theorem 2 and Corollary 1 are stated as functions of population parameters (SNR psi, N, p, q). The EM-GVA contraction (Lemma 1) and the n-free N^{-1/2} component follow from standard EM fixed-point analysis under the SNR condition, not from a quantity that has been fitted and then re-predicted. The only self-citations (Hong et al. 2021, Xiong et al. 2023) supply the external embeddings used as input features in the application; they do not establish the theorem or the model assumptions, so they are transfer-learning inputs rather than load-bearing circular evidence. The Discussion explicitly concedes the limitation that if external and target factor structures diverge substantially, rescaling eigenvalues is insufficient; that is an assumption/robustness limitation, not a circular reduction. The manuscript omits proofs of Lemma 1 and Theorem 2, which is a completeness concern for verification, but omission is not circularity. No quoted equation reduces to a fitted parameter or to a self-citation chain.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim depends on the correctness of the PALM model and the quality of the external pre-trained embeddings, which are domain assumptions rather than free parameters fitted in the paper. No new physical or conceptual entities are introduced.

free parameters (4)
  • q (latent embedding dimension)
    The number of latent factors is chosen a priori (e.g., by SVD truncation), not estimated by the model. The theory allows q to diverge.
  • K_B and K_M
    Positive constants in the parameter spaces Ω_B and Ω_M in eq (4). Their values are not specified but are assumed finite in the theory.
  • manual feature set X_man in MS application = 5 features
    Five EHR concepts were manually selected in Section 5 because they are known to relate to disability, a modeling choice that could influence results.
  • aggregation window = 6 months
    Visit feature counts are aggregated within a 6-month window before and after each visit, a choice made in the real data analysis.
assumptions (5)
  • domain assumption The PALM model (1) correctly specifies the joint distribution of X, U, Y: X|Z ~ Poisson(e^Z) with Z = V(B U_tilde_Y + W), W ~ N(0,Λ), Y|U ~ Bernoulli(expit(b^T U_tilde)).
    All derivations of the ELBO and error rates are based on this generative model. If real EHR counts are zero-inflated or overdispersed in ways not captured by Poisson log-normal, the theoretical guarantees may not hold.
  • domain assumption The pre-trained embedding matrix V is known, satisfies the orthonormality and incoherence conditions in Assumption 2(i), and spans the latent subspace of the target population.
    V is used directly as the factor loading space, bypassing estimation of a p×q matrix. If V is misaligned with the target cohort's disease structure, the model is misspecified.
  • domain assumption Labels are missing at random conditional on U: δ_i is independent of (X_i,Y_i) given U_i.
    Section 2.1 states the labeling process can depend on U_i. This assumption is needed for the observed-data likelihood to be valid in the semi-supervised setting.
  • standard math Technical Assumptions 1-3 hold: compactness of U, spectral bounds on Λ_0 and B_0, rate conditions among n,N,p,q, and the SNR condition ψ^2 > Cψ.
    These regularity conditions are used in the proofs of Theorems 1-2 and Lemma 1. They are standard in high-dimensional statistics but not proven in the paper.
  • domain assumption The external code embeddings from Xiong et al. (2023) and related sources accurately capture semantic and clinical structure of EHR concepts.
    In the MS application, V is constructed from these embeddings. The method's performance depends on their relevance and quality.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Semi-supervised Clustering Through Representation Learning of Large-scale EHR Data." pith.science (2026). https://pith.science/paper/VAI3IYGF

@misc{pith2026250520731,
  author       = {Pith},
  title        = {Pith review of: Semi-supervised Clustering Through Representation Learning of Large-scale EHR Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VAI3IYGF}},
  note         = {Machine review of arXiv:2505.20731}
}
read the original abstract

Electronic Health Records (EHR) offer rich real-world data for personalized medicine, providing insights into disease progression, treatment responses, and patient outcomes. However, their sparsity, heterogeneity, and high dimensionality make them difficult to model, while the lack of standardized ground truth further complicates predictive modeling. To address these challenges, we propose SCORE, a semi-supervised representation learning framework that captures multi-domain disease profiles through patient embeddings. SCORE employs a Poisson-Adapted Latent factor Mixture (PALM) Model with pre-trained code embeddings to characterize codified features and extract meaningful patient phenotypes and embeddings. To handle the computational challenges of large-scale data, it introduces a hybrid Expectation-Maximization (EM) and Gaussian Variational Approximation (GVA) algorithm, leveraging limited labeled data to refine estimates on a vast pool of unlabeled samples. We theoretically establish the convergence of this hybrid approach, quantify GVA errors, and derive SCORE's error rate under diverging embedding dimensions. Our analysis shows that incorporating unlabeled data enhances accuracy and reduces sensitivity to label scarcity. Extensive simulations confirm SCORE's superior finite-sample performance over existing methods. Finally, we apply SCORE to predict disability status for patients with multiple sclerosis (MS) using partially labeled EHR data, demonstrating that it produces more informative and predictive patient embeddings for multiple MS-related conditions compared to existing approaches.

Figures

Figures reproduced from arXiv: 2505.20731 by the authors.

Figure 1
Figure 1. Relative F-norm errors in estimating B and Λ across different methods. Shown are the relative errors ∥Bb − B0∥F /∥B0∥F and ∥Λb − Λ0∥F /∥Λ0∥F under varying values of N and n, with p = 400 and q = 20 held fixed. 21 [PITH_FULL_IMAGE:figures/full_fig_p021_1.png] view at source ↗
Figure 2
Figure 2. Relative F-norm errors in estimating B and Λ across different methods. Shown are the relative errors ∥Bb − B0∥F /∥B0∥F and ∥Λb − Λ0∥F /∥Λ0∥F under varying values of p and q, with N = 5000 and n = 100 held fixed [PITH_FULL_IMAGE:figures/full_fig_p022_2.png] view at source ↗
Figure 3
Figure 3. Mean cosine similarity between true and predicted patient embeddings across [PITH_FULL_IMAGE:figures/full_fig_p022_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Classification performance of different methods under various generative settings. [PITH_FULL_IMAGE:figures/full_fig_p023_4.png]
Figure 5
Figure 5. Figure 5: Classification performance with varying n for phenotyping of disability for MS patients at MGB. PDDS: trained and validated using disability status based on PDDS, PDDS-EDSS: trained using disability status based on PDDS and validated on EDSS-based labels. 25 [PITH_FUL…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Latent Factor Point Processes for Patient Representation in Electronic Health Records

    stat.ME 2025-08 reject novelty 5.0 of 10

    A latent factor point process model plus Fourier spectral embeddings gives new patient-level representations for EHR classification and clustering, but the stated theoretical guarantees contain a diverging error term.

Reference graph

Works this paper leans on

51 extracted references · 41 canonical work pages · cited by 1 Pith paper

  1. [1]

    The multivariate poisson-log normal distribution

    John Aitchison and CH Ho. The multivariate poisson-log normal distribution. Biometrika, 76 0 (4): 0 643--653, 1989

    work page 1989

  2. [2]

    Statistical guarantees for the em algorithm: From population to sample-based analysis

    Sivaraman Balakrishnan, Martin J Wainwright, and Bin Yu. Statistical guarantees for the em algorithm: From population to sample-based analysis. 2017

    work page 2017

  3. [3]

    Zero-inflation in the Multivariate Poisson Lognormal Family

    Bastien Batardi \`e re, Julien Chiquet, Fran c ois Gindraud, and Mahendra Mariadassou. Zero-inflation in the multivariate poisson lognormal family. arXiv preprint arXiv:2405.14711, 2024

    work page Pith review arXiv 2024

  4. [4]

    Representation learning: A review and new perspectives

    Yoshua Bengio, Aaron Courville, and Pascal Vincent. Representation learning: A review and new perspectives. IEEE transactions on pattern analysis and machine intelligence, 35 0 (8): 0 1798--1828, 2013

    2013

  5. [5]

    Asymptotic normality of maximum likelihood and its variational approximation for stochastic blockmodels

    Peter Bickel, David Choi, Xiangyu Chang, and Hai Zhang. Asymptotic normality of maximum likelihood and its variational approximation for stochastic blockmodels. 2013

    work page 2013

  6. [6]

    Variational inference: A review for statisticians

    David M Blei, Alp Kucukelbir, and Jon D McAuliffe. Variational inference: A review for statisticians. Journal of the American statistical Association, 112 0 (518): 0 859--877, 2017

    2017

  7. [7]

    Chime: Clustering of high-dimensional gaussian mixtures with em algorithm and its optimality

    T Tony Cai, Jing Ma, and Linjun Zhang. Chime: Clustering of high-dimensional gaussian mixtures with em algorithm and its optimality. The Annals of Statistics, 47 0 (3), 2019

    work page 2019

  8. [8]

    Robust principal component analysis? Journal of the ACM (JACM), 58 0 (3): 0 1--37, 2011

    Emmanuel J Cand \`e s, Xiaodong Li, Yi Ma, and John Wright. Robust principal component analysis? Journal of the ACM (JACM), 58 0 (3): 0 1--37, 2011

    work page 2011

Show all 51 references
  1. [9]

    Consistency of maximum-likelihood and variational estimators in the stochastic block model

    Alain Celisse, Jean-Jacques Daudin, and Laurent Pierre. Consistency of maximum-likelihood and variational estimators in the stochastic block model. 2012

    2012

  2. [10]

    Joint maximum likelihood estimation for high-dimensional exploratory item factor analysis

    Yunxiao Chen, Xiaoou Li, and Siliang Zhang. Joint maximum likelihood estimation for high-dimensional exploratory item factor analysis. Psychometrika, 84: 0 124--146, 2019

    2019

  3. [11]

    Variational inference for probabilistic poisson pca

    Julien Chiquet, Mahendra Mariadassou, and St \'e phane Robin. Variational inference for probabilistic poisson pca. The Annals of Applied Statistics, 12 0 (4): 0 2674--2698, 2018

    2018

  4. [12]

    Variational inference for sparse network reconstruction from count data

    Julien Chiquet, Stephane Robin, and Mahendra Mariadassou. Variational inference for sparse network reconstruction from count data. In International Conference on Machine Learning, pages 1162--1171. PMLR, 2019

    2019

  5. [13]

    Bert: Pre-training of deep bidirectional transformers for language understanding

    Jacob Devlin. Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805, 2018

    2018 arXiv

  6. [14]

    Bert: Pre-training of deep bidirectional transformers for language understanding

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human langu...

    2019

  7. [15]

    A markov chain monte carlo approach to confirmatory item factor analysis

    Michael C Edwards. A markov chain monte carlo approach to confirmatory item factor analysis. Psychometrika, 75 0 (3): 0 474--497, 2010

    2010

  8. [16]

    Factor augmented sparse throughput deep relu neural networks for high dimensional regression

    Jianqing Fan and Yihong Gu. Factor augmented sparse throughput deep relu neural networks for high dimensional regression. Journal of the American Statistical Association, 119 0 (548): 0 2680--2694, 2024

    2024

  9. [17]

    The patient-determined disease steps scale is not interchangeable with the expanded disease status scale in mild to moderate multiple sclerosis

    Yi Chao Foong, Daniel Merlo, Melissa Gresle, Chao Zhu, Katherine Buzzard, Jeannette Lechner-Scott, Michael Barnett, Bruce Taylor, Tomas Kalincik, Trevor Kilpatrick, et al. The patient-determined disease steps scale is not interchangeable with the expanded disease status scale ...

    2024

  10. [18]

    Theory of gaussian variational approximation for a poisson mixed model

    Peter Hall, John T Ormerod, and Matt P Wand. Theory of gaussian variational approximation for a poisson mixed model. Statistica Sinica, pages 369--389, 2011 a

    2011

  11. [19]

    Asymptotic normality and valid inference for gaussian variational approximation

    Peter Hall, Tung Pham, Matt P Wand, and Shen SJ Wang. Asymptotic normality and valid inference for gaussian variational approximation. 2011 b

    2011

  12. [20]

    Modern factor analysis

    Harry H Harman. Modern factor analysis. University of Chicago press, 1976

    1976

  13. [21]

    Clinical knowledge extraction via sparse embedding regression (keser) with multi-center large scale electronic health record data

    Chuan Hong, Everett Rush, Molei Liu, Doudou Zhou, Jiehuan Sun, Aaron Sonabend, Victor M Castro, Petra Schubert, Vidul A Panickan, Tianrun Cai, et al. Clinical knowledge extraction via sparse embedding regression (keser) with multi-center large scale electronic health record da...

    2021

  14. [22]

    Estimation of generalized linear latent variable models

    Philippe Huber, Elvezio Ronchetti, and Maria-Pia Victoria-Feser. Estimation of generalized linear latent variable models. Journal of the Royal Statistical Society Series B: Statistical Methodology, 66 0 (4): 0 893--908, 2004

    2004

  15. [23]

    An introduction to variational methods for graphical models

    Michael I Jordan, Zoubin Ghahramani, Tommi S Jaakkola, and Lawrence K Saul. An introduction to variational methods for graphical models. Machine learning, 37: 0 183--233, 1999

    1999

  16. [24]

    Em algorithm for mixed poisson and other discrete distributions

    Dimitris Karlis. Em algorithm for mixed poisson and other discrete distributions. ASTIN Bulletin: The Journal of the IAA, 35 0 (1): 0 3--24, 2005

    2005

  17. [25]

    Disability in multiple sclerosis: a reference for patients and clinicians

    Ilya Kister, Eric Chamot, Amber R Salter, Gary R Cutter, Tamar E Bacon, and Joseph Herbert. Disability in multiple sclerosis: a reference for patients and clinicians. Neurology, 80 0 (11): 0 1018--1024, 2013

    2013

  18. [26]

    Automatic variational inference in stan

    Alp Kucukelbir, Rajesh Ranganath, Andrew Gelman, and David Blei. Automatic variational inference in stan. Advances in neural information processing systems, 28, 2015

    2015

  19. [27]

    Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (edss)

    John F Kurtzke. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (edss). Neurology, 33 0 (11): 0 1444--1444, 1983

    1983

  20. [28]

    Deep representation learning of electronic health records to unlock patient stratification at scale

    Isotta Landi, Benjamin S Glicksberg, Hao-Chih Lee, Sarah Cherng, Giulia Landi, Matteo Danieletto, Joel T Dudley, Cesare Furlanello, and Riccardo Miotto. Deep representation learning of electronic health records to unlock patient stratification at scale. NPJ digital medicine, 3...

    2020

  21. [29]

    Validation of patient determined disease steps (pdds) scale scores in persons with multiple sclerosis

    Yvonne C Learmonth, Robert W Motl, Brian M Sandroff, John H Pula, and Diego Cadavid. Validation of patient determined disease steps (pdds) scale scores in persons with multiple sclerosis. BMC neurology, 13: 0 1--8, 2013

    2013

  22. [30]

    Biobert: a pre-trained biomedical language representation model for biomedical text mining

    Jinhyuk Lee, Wonjin Yoon, Sungdong Kim, Donghyeon Kim, Sunkyu Kim, Chan Ho So, and Jaewoo Kang. Biobert: a pre-trained biomedical language representation model for biomedical text mining. Bioinformatics, 36 0 (4): 0 1234--1240, 2020

    2020

  23. [31]

    Multisource representation learning for pediatric knowledge extraction from electronic health records

    Mengyan Li, Xiaoou Li, Kevin Pan, Alon Geva, Doris Yang, Sara Morini Sweet, Clara-Lea Bonzel, Vidul Ayakulangara Panickan, Xin Xiong, Kenneth Mandl, et al. Multisource representation learning for pediatric knowledge extraction from electronic health records. NPJ Digital Medici...

    2024

  24. [32]

    Efficient estimation of word representations in vector space

    Tomas Mikolov. Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781, 2013

    2013 arXiv

  25. [33]

    Deep patient: an unsupervised representation to predict the future of patients from the electronic health records

    Riccardo Miotto, Li Li, Brian A Kidd, and Joel T Dudley. Deep patient: an unsupervised representation to predict the future of patients from the electronic health records. Scientific reports, 6 0 (1): 0 1--10, 2016

    2016

  26. [34]

    Generalized latent trait models

    Irini Moustaki and Martin Knott. Generalized latent trait models. Psychometrika, 65: 0 391--411, 2000

    2000

  27. [35]

    The variational gaussian approximation revisited

    Manfred Opper and C \'e dric Archambeau. The variational gaussian approximation revisited. Neural computation, 21 0 (3): 0 786--792, 2009

    2009

  28. [36]

    A variational bayes approach to variable selection

    John T Ormerod, Chong You, and Samuel M \"u ller. A variational bayes approach to variable selection. 2017

    2017

  29. [37]

    Deep representation learning: Fundamentals, technologies, applications, and open challenges

    Amirreza Payandeh, Kourosh T Baghaei, Pooya Fayyazsanavi, Somayeh Bakhtiari Ramezani, Zhiqian Chen, and Shahram Rahimi. Deep representation learning: Fundamentals, technologies, applications, and open challenges. IEEE Access, 11: 0 137621--137659, 2023

    2023

  30. [38]

    Med-bert: pretrained contextualized embeddings on large-scale structured electronic health records for disease prediction

    Laila Rasmy, Yang Xiang, Ziqian Xie, Cui Tao, and Degui Zhi. Med-bert: pretrained contextualized embeddings on large-scale structured electronic health records for disease prediction. NPJ digital medicine, 4 0 (1): 0 86, 2021

    2021

  31. [39]

    High-dimensional maximum marginal likelihood item factor analysis by adaptive quadrature

    Stephen Schilling and R Darrell Bock. High-dimensional maximum marginal likelihood item factor analysis by adaptive quadrature. Psychometrika, 70: 0 533--555, 2005

    2005

  32. [40]

    A multivariate poisson-log normal mixture model for clustering transcriptome sequencing data

    Anjali Silva, Steven J Rothstein, Paul D McNicholas, and Sanjeena Subedi. A multivariate poisson-log normal mixture model for clustering transcriptome sequencing data. BMC bioinformatics, 20 0 (1): 0 1--11, 2019

    2019

  33. [41]

    Probabilistic principal component analysis

    Michael E Tipping and Christopher M Bishop. Probabilistic principal component analysis. Journal of the Royal Statistical Society Series B: Statistical Methodology, 61 0 (3): 0 611--622, 1999

    1999

  34. [42]

    Convergence properties of a general algorithm for calculating variational bayesian estimates for a normal mixture model

    DM Titterington and Bo Wang. Convergence properties of a general algorithm for calculating variational bayesian estimates for a normal mixture model. 2006

    2006

  35. [43]

    Asymptotic statistics, volume 3

    Aad W Van der Vaart. Asymptotic statistics, volume 3. Cambridge university press, 2000

    2000

  36. [44]

    Outcome measures in clinical trials for multiple sclerosis

    Caspar EP van Munster and Bernard MJ Uitdehaag. Outcome measures in clinical trials for multiple sclerosis. CNS drugs, 31 0 (3): 0 217--236, 2017

    2017

  37. [45]

    Graphical models, exponential families, and variational inference

    Martin J Wainwright, Michael I Jordan, et al. Graphical models, exponential families, and variational inference. Foundations and Trends in Machine Learning, 1 0 (1--2): 0 1--305, 2008

    2008

  38. [46]

    Knowledge-driven online multimodal automated phenotyping system

    Xin Xiong, Sara Morini Sweet, Molei Liu, Chuan Hong, Clara-Lea Bonzel, Vidul Ayakulangara Panickan, Doudou Zhou, Linshanshan Wang, Lauren Costa, Yuk-Lam Ho, et al. Knowledge-driven online multimodal automated phenotyping system. medRxiv, pages 2023--09, 2023

    2023

  39. [47]

    Yes, but did it work?: Evaluating variational inference

    Yuling Yao, Aki Vehtari, Daniel Simpson, and Andrew Gelman. Yes, but did it work?: Evaluating variational inference. In International Conference on Machine Learning, pages 5581--5590. PMLR, 2018

    2018

  40. [48]

    On variational bayes estimation and variational information criteria for linear regression models

    Chong You, John T Ormerod, and Samuel Mueller. On variational bayes estimation and variational information criteria for linear regression models. Australian & New Zealand Journal of Statistics, 56 0 (1): 0 73--87, 2014

    2014

  41. [49]

    Coder: Knowledge-infused cross-lingual medical term embedding for term normalization

    Zheng Yuan, Zhengyun Zhao, Haixia Sun, Jiao Li, Fei Wang, and Sheng Yu. Coder: Knowledge-infused cross-lingual medical term embedding for term normalization. Journal of biomedical informatics, 126: 0 103983, 2022

    2022

  42. [50]

    A note on exploratory item factor analysis by singular value decomposition

    Haoran Zhang, Yunxiao Chen, and Xiaoou Li. A note on exploratory item factor analysis by singular value decomposition. Psychometrika, 85 0 (2): 0 358--372, 2020

    2020

  43. [51]

    Multi-source learning via completion of block-wise overlapping noisy matrices

    Doudou Zhou, Tianxi Cai, and Junwei Lu. Multi-source learning via completion of block-wise overlapping noisy matrices. Journal of Machine Learning Research, 24 0 (221): 0 1--43, 2023

    2023

Pith tools

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