Pith. sign in

REVIEW 4 major objections 6 minor 30 references

Blood Glucose Level Prediction in Type 1 Diabetes Using Machine Learning

T0 review · 4 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read A voting regressor averaging MLP, LSTM, and GRU outputs achieves the best 30-minute-ahead blood glucose prediction RMSE of 22.50 mg/dL on the DiaTrend dataset, with different models leading in different glycemic ranges.

desk verdict A useful but flawed DiaTrend benchmark: the subject split is right, but the headline RMSE comparison is undermined by an ambiguous evaluation target and a missing baseline. read the letter →

arxiv 2502.00065 v1 pith:BJNUYQQL submitted 2025-01-30 q-bio.QM cs.LG

Soon Jynn Chu , Nalaka Amarasiri , Sandesh Giri , Priyata Kafle This is my paper · ORCID
classification q-bio.QMcs.LG
keywords bloodglucosepredictiontype1diabetescontinuousmonitoringvotingregressordeepreinforcementlearningDiaTrenddataset30-minutehorizon
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper asks which machine-learning approach best predicts blood glucose 30 minutes ahead from continuous glucose monitor (CGM) data, using a recent public dataset. On the DiaTrend dataset, 15 models were trained—classical regressors (SVR, random forest, LightGBM), deep networks (MLP, LSTM, GRU), deep reinforcement learning agents (DDPG, TD3, SAC), and voting/stacking ensembles of each family—with training and test sets drawn from different patients. The central claim is that a voting regressor averaging MLP, LSTM, and GRU outputs achieves the best overall RMSE of 22.50 mg/dL, edging out the corresponding stacking model by 0.03 mg/dL. When errors are split by glycemic range, the best model differs by range: LSTM and the stacking model lead in normoglycemia, the voting ensemble leads in hyperglycemia, and DDPG leads in hypoglycemia. A sympathetic reading is that ensembling diverse neural architectures is a strong default for short-term glucose prediction, and that the rare, dangerous low-glucose range needs its own specialized model.

What carries the argument

The central object is the voting regressor V2, which averages the outputs of three neural sequence models—an MLP, an LSTM, and a GRU—each trained to map a six-point CGM input window to the next six glucose readings. The supporting pipeline slices each patient's continuous glucose stream into 250,559 six-point windows, normalizes values to [-1, 1], and enforces a subject-disjoint split (11 training subjects, 6 test subjects) to avoid patient leakage. The other key device is the glycemic-condition breakdown, which re-aggregates the same squared errors into normoglycemia, hyperglycemia, and hypoglycemia bins to show that the overall best model is not the best in every range.

What would settle it

Recompute the comparison with a patient-level block bootstrap so that all windows from one subject stay together; if the RMSE gap between V2 (22.50) and S2 (22.53) reverses or the confidence intervals overlap substantially, the claimed overall best-model ranking is not established. A second check is to recount hypoglycemic events in the test set and recompute DDPG's RMSE there after removing the most extreme outlier events.

Watch

Extended reading notes

Core claim

The paper's central claim is that Model V2, a voting regressor that averages the pointwise predictions of an MLP, an LSTM, and a GRU, is the best overall model for 30-minute-ahead blood glucose prediction on the DiaTrend dataset, reaching an RMSE of 22.50 mg/dL on six held-out subjects. The same three networks in a stacking configuration, S2, posts the best MAE (16.29 mg/dL) and MAPE (10.40%) and trails V2 by only 0.03 mg/dL in RMSE. Stratifying by glycemic condition, LSTM and S2 tie for the best normoglycemic RMSE at 19.03 mg/dL, V2 is best in hyperglycemia at 26.02 mg/dL, and DDPG is best in hypoglycemia at 28.25 mg/dL, beating the runner-up MLP by 0.79 mg/dL. The paper presents these results as demonstrating that different models lead in different glycemic ranges, with the neural voting ensemble the strongest overall performer.

Load-bearing premise

The paper treats each overlapping six-point window cut from a patient's continuous glucose stream as an independent sample, even though consecutive windows share five of six points, making 250,559 examples far fewer independent observations than they appear; if that independence fails, the reported standard deviations and the near-tie ranking are unreliable.

Editorial extensions

If this is right

  • If V2's advantage is real, averaging the outputs of an MLP, an LSTM, and a GRU is a solid, simple default for 30-minute CGM-based glucose forecasting.
  • The 0.03 mg/dL gap between voting and stacking suggests the meta-learner choice matters little; either ensemble of the same three networks is essentially equivalent.
  • DDPG's lead in hypoglycemia suggests that a state-dependent routing policy—using a DRL agent when glucose is low—could improve overall performance without sacrificing the ensemble's edge elsewhere.
  • The subject-disjoint split means these numbers are a fair estimate of person-level generalization on DiaTrend, giving other researchers a benchmark to beat.
  • Including basal and bolus insulin data plus meal records, which the paper lists as future work, is the most direct next step beyond these CGM-only results.

Reading between the lines

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

  • A subject-level bootstrap would probably show that V2 and S2 are statistically indistinguishable, so the 22.50 versus 22.53 headline difference is likely within noise.
  • Because hypoglycemia is only about 1% of the data, DDPG's RMSE lead there rests on a handful of low-glucose episodes; its high MAPE of roughly 38% warns against treating it as a reliable hypo-alarm model.
  • The consistent right-shift in the prediction plots suggests a systematic lag; a simple lag-correction or a longer input context might improve all models more than changing the regressor.
  • A natural test is to replace the three neural bases with a modern sequence model such as a transformer or a temporal convolutional network to see whether the ensemble advantage persists.
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 compares 15 machine-learning models for 30-minute-ahead blood glucose prediction on the DiaTrend dataset, using a subject-level split into 11 training and 6 test subjects. The models include traditional regressors (SVR, RF, LGB), deep neural networks (MLP, LSTM, GRU), deep reinforcement learners (DDPG, TD3, SAC), and voting/stacking ensembles built from these base estimators. The authors report that voting ensemble V2 (MLP+LSTM+GRU) achieves the lowest overall RMSE of 22.50 mg/dL, while LSTM and S2 lead in normoglycemia and DDPG leads in hypoglycemia with RMSE 28.25. The paper argues these results demonstrate the performance of various models across different glycemic conditions, and it provides source code.

Significance. If the central evaluation issues were resolved, this would be a useful subject-level benchmark for 30-minute-ahead CGM-based glucose prediction on the DiaTrend dataset. The paper has several strengths: it uses a subject-level train/test split (avoiding patient overlap between train and test), compares a wide range of model families including DRL and ensembles, reports per-glycemic-condition results, and makes code publicly available. However, the current manuscript does not support its headline ranking claim because the target definition for the DNN/DRL models is ambiguous, there is no baseline or significance testing, and the reported standard deviations are computed under an independence assumption that is violated by the overlapping-window construction.

major comments (4)
  1. [§3.2, Tables 3-4] The manuscript never specifies whether the MAE, MAPE, and RMSE values in Tables 3 and 4 are computed from all six elements of the output sequence or only from the final, 30-minute-ahead element. Section 3.2 states that traditional ML models use the last value of the output sequence as the target, while DNNs and DRL agents generate an output sequence of the same length as the input; the evaluation section then reports a single set of metrics for every model. If the metrics average over all six output elements, then Table 3 reports a mixture of 5-, 10-, 15-, 20-, 25-, and 30-minute horizons, and the headline comparison (V2 at 22.50 vs S2 at 22.53) is not a 30-minute-ahead comparison. If only the last element is used, the text should say so explicitly, because the architecture descriptions and the phrase "generate an output sequence" imply otherwise. This ambiguity directly affects the paper's central claim and must be resolved before the results can be interpreted.
  2. [§3.1] The 250,559 windows are constructed by sliding a one-hour window over continuous CGM streams, so consecutive windows from the same subject share five of six input points and five of six output points. Treating every window as an independent sample therefore overstates the effective sample size by a large factor, and the per-subject standard deviations reported in Tables 3 and 4 do not capture the true uncertainty of the model comparisons. In particular, the difference between V2 (22.50) and S2 (22.53) is two orders of magnitude smaller than the reported standard deviations, so without a proper accounting for the autocorrelation in the data the ranking of the top models is not reliable. The authors should either use non-overlapping windows or report cluster-robust or subject-level paired statistics.
  3. [§4] No baseline or significance test is reported. The claim that V2 is "the best-performing model overall" rests on comparing 15 models on a single 6-subject test set, with no persistence or naive baseline and no paired significance test (for example, a paired test over the 6 subjects or a bootstrap over subjects). Given that the differences between the top models are within the reported standard deviations and there is no correction for multiple comparisons, the current evidence does not support the central ranking claim. Adding a baseline and appropriate hypothesis tests, or at least acknowledging the absence of statistical significance, is necessary.
  4. [§3.2, §4] The model comparison appears to select the best model on the same test set used for the final comparison, with no separate validation set or nested procedure. Since hyperparameters, training epochs, and the choice of "best model" are all determined using the same 6 subjects, the reported winners are at risk of being overfit to the test set. The authors should clarify whether any validation split was used for early stopping and model selection, and should discuss the multiple-comparisons issue when claiming a single best model.
minor comments (6)
  1. [Figures 6-7] The legends in Figures 6 and 7 list V4 and S4, but Table 2 defines only V1-V3 and S1-S3; the legends should be corrected.
  2. [Figures 3-7] The captions contain the typo "Grount Truth" for "Ground Truth".
  3. [§3.2] "Multi-Layer Perception" should be "Multi-Layer Perceptron".
  4. [§3.1] The subject-level split is a strength, but the phrase "randomly split" without a seed or a description of the random process makes the exact split non-reproducible; please provide the seed or a reproducible split procedure.
  5. [§3.2, Eq. (2)] Equation (2) defines a reward function as 1 - RMSE/(max y - min y), which is not the standard normalized RMSE; consider using a different name or providing a citation for this convention.
  6. [§1] The sentence "we aim to identify the most effective methods for accurate insulin level prediction" appears to be a wording error; the paper predicts blood glucose levels, not insulin levels.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: empirical benchmark with held-out subjects; all target values derive from CGM data, not from fitted parameters.

full rationale

This is an empirical benchmarking study rather than a derivation chain. The 30-minute-ahead prediction target is defined directly from the CGM record in Section 3.1, and each model is trained on 11 subjects and evaluated on 6 held-out subjects, so no fitted parameter is being renamed as a prediction. The normalization bounds and hyperparameters are explicit modeling choices, not quantities that are later reported as results. The only self-citation identified is an author's PhD thesis cited in a general related-work sentence about reinforcement learning applications; it is not load-bearing and does not support any of the paper's measured outcomes. The reviewer's noted horizon ambiguity and the overlapping-window non-independence are validity and reporting concerns, not cases where a reported quantity is equivalent to an input by construction. No equation or claim in the paper reduces to its own inputs, and no load-bearing self-citation chain is present. Therefore the circularity score is 0.

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

The paper introduces no new physical or biological entities. It rests on domain assumptions about CGM-only data sufficiency, unbiased missing-value removal, representativeness of a single subject split, and independence of overlapping windows. The central benchmark depends on several hand-chosen constants and hyperparameters, but none are presented as derived predictions.

free parameters (4)
  • Normalization bounds = min_x = 20, max_x = 420 mg/dL
    Hand-chosen constants in Eq. 1; they scale inputs and targets and affect the DRL reward, but they are not fitted to data.
  • Input window length and prediction horizon = 30 minutes input (6 points), 30 minutes output (6 points)
    Problem definition choice made by the authors; the central task and all results depend on these values.
  • DNN hyperparameters = MLP hidden 256, LSTM hidden 75, GRU hidden 86, dropout 0.2, learning rate 0.01, 100 epochs, batch size 32
    Hand-set in Table 1 and not tuned on a validation set; they directly affect the model comparison.
  • DRL training budget and learning rate = 10 epochs, learning rate 1e-4
    Ad hoc training budget for DDPG, TD3, and SAC; likely insufficient for RL convergence and not justified in the paper.
assumptions (5)
  • domain assumption Six CGM readings over 30 minutes contain enough information to predict glucose 30 minutes ahead.
    Section 3.1 defines the task using only the CGM feature; no insulin, meal, or activity data are used.
  • domain assumption Removing samples with missing values does not bias the training or test distribution.
    Section 3.1 states missing samples were removed because sufficient data remained; if missingness is correlated with glucose values, this is selection bias.
  • domain assumption The six test subjects are representative of the T1D population and a single split is sufficient.
    Section 3.1 uses one 11/6 subject split from 17 patients in one study; no cross-validation or repeated splits are performed.
  • domain assumption Overlapping six-point windows from the same subject can be treated as independent samples.
    Section 3.1 builds all windows from continuous CGM streams; consecutive windows share five of six points, so the effective sample size is far below 250,559.
  • domain assumption Default hyperparameters in scikit-learn, PyTorch, and d3rlpy are reasonable for a fair comparison.
    Section 3.2 keeps many hyperparameters at defaults; this is a modeling assumption that affects the ranking.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Blood Glucose Level Prediction in Type 1 Diabetes Using Machine Learning." pith.science (2026). https://pith.science/paper/BJNUYQQL

@misc{pith2026250200065,
  author       = {Pith},
  title        = {Pith review of: Blood Glucose Level Prediction in Type 1 Diabetes Using Machine Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BJNUYQQL}},
  note         = {Machine review of arXiv:2502.00065}
}
read the original abstract

Type 1 Diabetes is a chronic autoimmune condition in which the immune system attacks and destroys insulin-producing beta cells in the pancreas, resulting in little to no insulin production. Insulin helps glucose in your blood enter your muscle, fat, and liver cells so they can use it for energy or store it for later use. If insulin is insufficient, it causes sugar to build up in the blood and leads to serious health problems. People with Type 1 Diabetes need synthetic insulin every day. In diabetes management, continuous glucose monitoring is an important feature that provides near real-time blood glucose data. It is useful in deciding the synthetic insulin dose. In this research work, we used machine learning tools, deep neural networks, deep reinforcement learning, and voting and stacking regressors to predict blood glucose levels at 30-min time intervals using the latest DiaTrend dataset. Predicting blood glucose levels is useful in better diabetes management systems. The trained models were compared using several evaluation metrics. Our evaluation results demonstrate the performance of various models across different glycemic conditions for blood glucose prediction. The source codes of this work can be found in: https://github.com/soon-jynn-chu/t1d_bg_prediction

Figures

Figures reproduced from arXiv: 2502.00065 by the authors.

Figure 1
Figure 1. Distribution for Each Condition for Train and Test Set [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. DNN Model Architecture Illustration All DNNs were implemented and trained using PyTorch (v2.4.0). The train￾ing hyperparameters are shown in [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. , 4, 5, 6, and 7 illustrate the prediction of each method against the ground truth, plotted half a day for Subject 37. All models show a similar prediction trend as the ground truth, but a slight shift to the right. 09-18 12 09-18 14 09-18 16 09-18 18 09-18 20 09-18 22 09-19 00 Date 100 150 200 250 300 Blood Glucose Level Grount Truth SVR RF LGB Hyperglycemia Hypoglycemia [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: RL methods prediction visual￾ization 09-18 12 09-18 14 09-18 16 09-18 18 09-18 20 09-18 22 09-19 00 Date 100 150 200 250 300 Blood Glucose Level Grount Truth V1 V2 V3 V4 Hyperglycemia Hypoglycemia [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]
Figure 7
Figure 7. Figure 7: Stack regressor methods prediction visualiza [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 29 canonical work pages

  1. [1]

    Type 1 diabetes

    Mark A Atkinson, George S Eisenbarth, and Aaron W Michels. Type 1 diabetes. The lancet, 383(9911):69–82, 2014

    work page 2014

  2. [2]

    Inpatient management of adults and children with type 1 diabetes.Endocrinology and Metabolism Clinics, 39(3):595–608, 2010

    David M Tridgell, Angela H Tridgell, and Irl B Hirsch. Inpatient management of adults and children with type 1 diabetes.Endocrinology and Metabolism Clinics, 39(3):595–608, 2010

    work page 2010

  3. [3]

    Epidemiologyoftype1diabetes

    David M Maahs, Nancy A West, Jean M Lawrence, and Elizabeth J Mayer-Davis. Epidemiologyoftype1diabetes. Endocrinology and Metabolism Clinics,39(3):481– 497, 2010

    work page 2010

  4. [4]

    The management of type 1 diabetes in adults

    Richard IG Holt, J Hans DeVries, Amy Hess-Fischl, Irl B Hirsch, M Sue Kirkman, Tomasz Klupa, Barbara Ludwig, Kirsten Nørgaard, Jeremy Pettus, Eric Renard, et al. The management of type 1 diabetes in adults. a consensus report by the american diabetes association (ada) and the european association for the study of diabetes (easd). Diabetes care, 44(11):258...

    work page 2021

  5. [5]

    R Livingstone, JG Boyle, and JR Petrie. How tightly controlled do fluctuations in blood glucose levels need to be to reduce the risk of developing complications in people with type 1 diabetes?Diabetic Medicine, 37(4):513–521, 2020

    work page 2020

  6. [6]

    Standards of care in diabetes—2023.Diabetes care, 46:S1– S267, 2023

    Diabetes Care et al. Standards of care in diabetes—2023.Diabetes care, 46:S1– S267, 2023

    work page 2023

  7. [7]

    Glycemic variability: Measurement, target, impact on complications of diabetes and does it really matter?Journal of Diabetes Investigation, 15(1):5–14, 2024

    Yifei Mo, Jingyi Lu, and Jian Zhou. Glycemic variability: Measurement, target, impact on complications of diabetes and does it really matter?Journal of Diabetes Investigation, 15(1):5–14, 2024

    work page 2024

  8. [8]

    Continuousglucosemonitor- ingdevices:Abriefpresentation

    Doina Andrada Mihai, Diana Simona Stefan, Daniela Stegaru, Georgiana Elena Bernea, Ileana Adela Vacaroiu, Toma Papacocea, Mircea Ovidiu Denis Lupus,oru, AdrianaElenaNica,OvidiuStiru,DorinDragos,etal. Continuousglucosemonitor- ingdevices:Abriefpresentation. Experimental and therapeutic medicine,23(2):1–6, 2022

    work page 2022

Show all 30 references
  1. [9]

    Clinical implications of real-time and intermittently scanned continuous glucose monitor- ing

    Steven V Edelman, Nicholas B Argento, Jeremy Pettus, and Irl B Hirsch. Clinical implications of real-time and intermittently scanned continuous glucose monitor- ing. Diabetes Care, 41(11):2265–2274, 2018. 14 Chu et al

    2018

  2. [10]

    Reproducibil- ity of continuous glucose monitoring results under real-life conditions in an adult population: a functional data analysis.Scientific Reports, 13(1):13987, 2023

    Marcos Matabuena, Marcos Pazos-Couselo, Manuela Alonso-Sampedro, Carmen Fernández-Merino, Arturo González-Quintela, and Francisco Gude. Reproducibil- ity of continuous glucose monitoring results under real-life conditions in an adult population: a functional data analysis.Scie...

    2023

  3. [11]

    Using continuous glucose monitoring in clinical practice.Clinical diabetes: a publication of the American Diabetes Association, 38(5):429, 2020

    Eden M Miller. Using continuous glucose monitoring in clinical practice.Clinical diabetes: a publication of the American Diabetes Association, 38(5):429, 2020

    2020

  4. [12]

    Be- yond a1c: exploring continuous glucose monitoring metrics in managing diabetes

    Jared G Friedman, Kasey Coyne, Grazia Aleppo, and Emily D Szmuilowicz. Be- yond a1c: exploring continuous glucose monitoring metrics in managing diabetes. Endocrine Connections, 12(7), 2023

    2023

  5. [13]

    Diatrend: A dataset from advanced diabetes technology to enable development of novel analytic solutions.Scientific Data, 10(1):556, 2023

    Temiloluwa Prioleau, Abigail Bartolome, Richard Comi, and Catherine Stanger. Diatrend: A dataset from advanced diabetes technology to enable development of novel analytic solutions.Scientific Data, 10(1):556, 2023

    2023

  6. [14]

    Predicting adverse events for patients with type-1 diabetes via self-supervised learning

    Xinzhe Zheng, Sijie Ji, and Chenshu Wu. Predicting adverse events for patients with type-1 diabetes via self-supervised learning. In ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 1526–1530. IEEE, 2024

    2024

  7. [15]

    T1diabetesgranada: a longitudinal multi-modal dataset of type 1 diabetes mellitus

    Ciro Rodriguez-Leon, Maria Dolores Aviles-Perez, Oresti Banos, Miguel Quesada- Charneco, Pablo J Lopez-Ibarra Lozano, Claudia Villalonga, and Manuel Munoz- Torres. T1diabetesgranada: a longitudinal multi-modal dataset of type 1 diabetes mellitus. Scientific Data, 10(1):916, 2023

    2023

  8. [16]

    Mealtime prediction using wearable insulin pump data to support diabetes management.Scientific Reports, 14(1):21013, 2024

    Baiying Lu, Yanjun Cui, Prajakta Belsare, Catherine Stanger, Xia Zhou, and Temiloluwa Prioleau. Mealtime prediction using wearable insulin pump data to support diabetes management.Scientific Reports, 14(1):21013, 2024

    2024

  9. [17]

    SawsanKurdi,AhmadAlamer,HaythamWali,AishaFBadr,MerriLPendergrass, NehadAhmed,IvoAbraham,andMaryamTFazel. Proof-of-conceptstudyofusing supervised machine learning algorithms to predict self-care and glycemic control in type 1 diabetes patients on insulin pump therapy.Endocrine ...

    2023

  10. [18]

    Universal Omni-Wheeled Mobile Robot Path Planning Using Reinforcement Learning Enhanced With Physics Based Modeling and Control

    Nalaka Amarasiri. Universal Omni-Wheeled Mobile Robot Path Planning Using Reinforcement Learning Enhanced With Physics Based Modeling and Control. PhD thesis, University of Louisiana at Lafayette, 2024

    2024

  11. [19]

    Adaptive voltage and frequency regulation for secondary control via reinforcement learning for islanded microgrids

    Kouhyar Sheida, Mohammad Seyedi, and Farzad Ferdowsi. Adaptive voltage and frequency regulation for secondary control via reinforcement learning for islanded microgrids. In 2024 IEEE Texas Power and Energy Conference (TPEC), pages 1–6. IEEE, 2024

    2024

  12. [20]

    Resilient reinforcement learning for voltage control in an islanded dc microgrid integrating data-driven piezoelectric

    Kouhyar Sheida, Mohammad Seyedi, Muhammad Ali Afridi, Farzad Ferdowsi, Mo- hammad J Khattak, Vijaya K Gopu, and Tyson Rupnow. Resilient reinforcement learning for voltage control in an islanded dc microgrid integrating data-driven piezoelectric. Machines, 12(10):694, 2024

    2024

  13. [21]

    A multimodal inter- mediate fusion network with manifold learning for stress detection.arXiv preprint arXiv:2403.08077, 2024

    Morteza Bodaghi, Majid Hosseini, and Raju Gottumukkala. A multimodal inter- mediate fusion network with manifold learning for stress detection.arXiv preprint arXiv:2403.08077, 2024

    2024 arXiv

  14. [22]

    Adaptive attention-based lstm framework for multimodal driver drowsiness prediction

    Morteza Bodaghi. Adaptive attention-based lstm framework for multimodal driver drowsiness prediction. Master’s thesis, University of Louisiana at Lafayette, 2024

    2024

  15. [23]

    Multimodal stress detection using facial landmarks and bio- metric signals

    Majid Hosseini, Morteza Bodaghi, Ravi Teja Bhupatiraju, Anthony Maida, and Raju Gottumukkala. Multimodal stress detection using facial landmarks and bio- metric signals. arXiv preprint arXiv:2311.03606, 2023

    2023 arXiv

  16. [24]

    Basal glucose control in type 1 diabetes using deep reinforcement learning: An in silico validation.IEEE Journal of Biomedical and Health Informatics, 25(4):1223–1232, 2020

    Taiyu Zhu, Kezhi Li, Pau Herrero, and Pantelis Georgiou. Basal glucose control in type 1 diabetes using deep reinforcement learning: An in silico validation.IEEE Journal of Biomedical and Health Informatics, 25(4):1223–1232, 2020. Blood Glucose Prediction using ML 15

    2020

  17. [25]

    A reinforcement learning–based method for management of type 1 diabetes: exploratory study.JMIR diabetes, 4(3):e12905, 2019

    Mahsa Oroojeni Mohammad Javad, Stephen Olusegun Agboola, Kamal Jethwani, Abe Zeid, Sagar Kamarthi, et al. A reinforcement learning–based method for management of type 1 diabetes: exploratory study.JMIR diabetes, 4(3):e12905, 2019

    2019

  18. [26]

    DiaTrend: A dataset from advanced diabetes technology to enable development of novel analytic solutions.Scientific Data, 10(1):556, 2023

    Temiloluwa Prioleau, Abigail Bartolome, Richard Comi, and Catherine Stanger. DiaTrend: A dataset from advanced diabetes technology to enable development of novel analytic solutions.Scientific Data, 10(1):556, 2023

    2023

  19. [27]

    Lillicrap, Jonathan J

    Timothy P. Lillicrap, Jonathan J. Hunt, Alexander Pritzel, Nicolas Heess, Tom Erez, Yuval Tassa, David Silver, and Daan Wierstra. Continuous control with deep reinforcement learning, July 2019

    2019

  20. [28]

    Addressing Function Approx- imation Error in Actor-Critic Methods, October 2018

    Scott Fujimoto, Herke van Hoof, and David Meger. Addressing Function Approx- imation Error in Actor-Critic Methods, October 2018

    2018

  21. [29]

    Soft Actor- Critic:Off-PolicyMaximumEntropy DeepReinforcement Learningwith aStochas- tic Actor, August 2018

    Tuomas Haarnoja, Aurick Zhou, Pieter Abbeel, and Sergey Levine. Soft Actor- Critic:Off-PolicyMaximumEntropy DeepReinforcement Learningwith aStochas- tic Actor, August 2018

    2018

  22. [30]

    D3rlpy: An offline deep reinforcement learning library

    Takuma Seno and Michita Imai. D3rlpy: An offline deep reinforcement learning library. Journal of Machine Learning Research, 23(315):1–20, 2022

    2022

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.