Pith. sign in

REVIEW 3 major objections 6 minor 43 references

A Pre-Trained Graph-Based Model for Adaptive Sequencing of Educational Documents

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper claims that pre-training a recommender on sequential course corpora makes learning-path personalization on a new adaptive corpus far more sample-efficient, especially at the start of fine-tuning.

desk verdict A promising transfer-learning idea that is undermined by an internal inconsistency in the transition function; the paper needs a code check and a corrected Eq. 3 before its empirical claims can be trusted. read the letter →

arxiv 2411.11520 v1 pith:WQM52TON submitted 2024-11-18 cs.AI cs.CYcs.LG

Jean Vassoyan (CB) , Anan Schütt (UNIA) , Jill-Jênn Vie (SODA) , Arun-Balajiee Lekshmi-Narayanan (PITT) , Elisabeth André (UNIA) , Nicolas Vayatis (CB) This is my paper · ORCID
classification cs.AIcs.CYcs.LG
keywords adaptivelearningpathpersonalizationreinforcementgraphneuralnetworkstransferpre-trainingrecommendersystemsampleefficiency
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 claims that a recommender system for learning path personalization can be pre-trained on ordinary sequential courses, the kind that make up most MOOC content, and then fine-tuned on a new, more flexible "graph" corpus with far fewer student interactions than training from scratch. This matters because collecting real student interaction data is expensive and expert tagging of knowledge components is rarely available. The proposed system needs neither: it extracts keywords from raw course materials with an LLM and learns from simulated students during pre-training. In experiments with a 22-document adaptive-learning corpus and 50 fine-tuning students, the pre-trained model outperformed all baselines in all three prior-knowledge scenarios, with the largest advantage at the start of fine-tuning.

What carries the argument

The central object is a two-part recommender system $F = \psi \circ \phi$: an encoder $\phi$ turns the student's feedback history into a keyword-vector knowledge state, and a policy $\psi$ scores candidate documents. Both are graph neural networks using the graph-transformer operator on a bipartite document\u2013keyword graph, so the parameter count does not depend on corpus size. Pre-training has two stages: supervised imitation of an oracle that knows the prerequisite graph, then REINFORCE on 14 sequential corpora. This pre-training is what gives the model a warm start when it is later fine-tuned on a target graph corpus.

What would settle it

Run the exact same pre-training and fine-tuning protocol with the 22-document graph corpus but replace the simulated students with human learners, or with a simulator fitted to human interaction logs; if the pre-trained model's learning-gain advantage over a from-scratch model does not appear within the first 50 learners, the central claim is falsified.

Watch

Extended reading notes

Core claim

The paper's central claim is that pre-training a graph-based recommender system on sequential educational corpora transfers to a new adaptive-learning corpus and yields substantially better sample efficiency than training from scratch. In the experiments, the pre-trained model outperformed all baselines in all three tested prior-knowledge scenarios: no prior knowledge, decreasing-exponential prior knowledge, and uniform prior knowledge. The gaps are largest at the start of fine-tuning, and final average learning gains per student after 50 students reach 24.81, 22.62, and 13.33 respectively, versus best baselines of 18.63, 16.28, and 7.51. The paper presents this as the first pre-trained recommender system for learning path personalization, operating without expert annotation.

Load-bearing premise

The entire evaluation assumes the simulated student dynamics, learning a knowledge component only when all prerequisites are mastered and otherwise learning nothing, match real students closely enough that the measured transfer gains would appear with human learners.

Editorial extensions

If this is right

  • A new adaptive-learning corpus can be bootstrapped from raw course materials alone, with no expert tagging of knowledge components or prerequisite graphs.
  • The largest benefit appears in the small-data regime: with only 50 simulated students, the pre-trained model's final learning gains exceed every baseline in all three prior-knowledge scenarios.
  • Pre-training on a zero-prior-knowledge population transfers to populations with decreasing-exponential and uniform prior knowledge, so the learned structure is not tied to the training population's distribution.
  • The reinforcement-learning stage of pre-training is necessary for the warm start, since pre-training by imitation alone or by next-feedback prediction yields much weaker early fine-tuning performance.
  • Because the model represents documents and students through keyword embeddings rather than corpus-specific identifiers, the same architecture applies to new corpora of different sizes.

Reading between the lines

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

  • A testable extension is to vary keyword overlap between source and target corpora; if transfer is driven by shared conceptual vocabulary, performance should track that overlap.
  • A human-subject replication with the same 22-document corpus would show whether the all-or-nothing prerequisite learning rule used in simulation overstates the warm-start benefit.
  • The low variance of the pre-trained model suggests the supervised imitation stage stabilizes the policy; one could test whether imitation plus any on-policy learner, not just REINFORCE, reproduces the transfer.
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 / 6 minor

Summary. This paper formalizes learning path personalization as a collection of POMDPs over randomly generated student-corpus graphs, partitions corpora into sequential (source) and graph (target) tasks, and proposes a GNN-based recommender pre-trained on 14 real MOOCs' sequential corpora (supervised oracle imitation plus REINFORCE) and then fine-tuned on a hand-designed 22-document graph corpus with simulated students. Across three prior-knowledge scenarios, the pre-trained model is reported to outperform the no-pre-training GNN baseline, a contextual bandit, and a PPO baseline, with the largest gain at the beginning of fine-tuning.

Significance. If the reported transfer effect is reproducible, the paper makes a useful step toward data-efficient adaptive learning and is, to my knowledge, one of the first pre-trained recommender systems for this setting. The work is transparent in several ways: the POMDP formalization is explicit, a code link is provided, results are averaged over 30 seeds with bootstrap confidence intervals, and Appendix E ablates the pre-training objective. However, the central evidence is currently compromised by an inconsistency in the core transition function (Eq. 3 vs. Eq. 6), and the evaluation is confined to one hand-crafted target corpus with simulated students generated under the same model family that the pre-training assumes. These issues must be resolved before the transfer claim can be accepted.

major comments (3)
  1. [Section 2.2, Eq. (3); Appendix A.2, Eq. (6)] Equation (3) is not a valid probability transition. For a knowledge component ki in d-> with ms(d<-)=1 and xi=1, the first line evaluates to 1*(1-1)+0=0 for x'i=1, and the second line gives 0 for x'i=0, so the distribution over x'i sums to 0. Appendix A.2's Eq. (6), which is claimed to be equivalent, gives probability 1 to x'i=1 in this case. Because every episode in the fine-tuning experiments (Section 5.4, Figure 4, Table 1) is simulated with this transition, the reported returns depend on whether the code implements Eq. (3), Eq. (6), or a third rule. The authors must correct Eq. (3) and explicitly verify it against the released code; if the code followed Eq. (3), all simulation-based results in Section 5 need to be re-run. As written, the paper is internally inconsistent and the central empirical claim cannot be checked.
  2. [Section 5.2 and 5.4, Appendix D.1] The fine-tuning reward is explicitly changed from Eq. (4) to a weighted sum, with KC values 1, 2, and 3 in Appendix D.1, yet Section 5.4 calls the reported quantity 'learning gains per student' and equates it with the undiscounted episodic return. These are not the same object. Please state the exact fine-tuning reward in the main text, report raw learning gains (number of newly acquired KCs) alongside the weighted return, or relabel all figures and tables as weighted returns; otherwise readers cannot interpret the magnitude of the reported values.
  3. [Section 5.1 and 5.4] The generalization claim is supported by a single target corpus, and the simulated students in that corpus are produced by the same transition and observation functions (Eqs. 1-4) that define the pre-training environment. This makes the warm-start benefit a within-model result. To support the claim of transfer to 'a range of adaptive learning scenarios,' please add at least one additional target corpus with a different structure, or a robustness study in which the student simulator is perturbed (e.g., noisy learning, forgetting, stochastic observations), and temper the conclusion accordingly if such experiments are not available.
minor comments (6)
  1. [Appendix A.2, Eq. (6)] The symbol sigma is used in Eq. (6) but never defined; from context it appears to denote the indicator that at least one prerequisite is unknown. Please define it explicitly.
  2. [Section 5.1 and Appendix D.1] It is not stated how many KCs each document teaches and how the 11-step episode limit relates to the 22 documents; please clarify this relation so the reported returns can be interpreted.
  3. [Section 3 and Appendix B] The symbol E denotes both the edge set Eprereq union Epref and the embedding matrix; please use distinct symbols to avoid ambiguity.
  4. [Section 4.2] The sentence 'we have trained the RL agent on approximately 25k steps' would benefit from stating the stopping criterion and whether the final checkpoint or the best checkpoint was used for fine-tuning.
  5. [Section 7] The word 'developement' should be 'development'.
  6. [Appendix D.2 and Table 3] The discount factor is 0 during fine-tuning but 0.7 during pre-training; a sentence explaining this choice would improve reproducibility.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: the transfer gain is an empirical result measured on held-out target corpora; only minor non-load-bearing self-citations appear, while the Eq. 3 / Eq. 6 inconsistency is a correctness issue, not circularity.

full rationale

The central claim—that pre-training on sequential corpora improves sample efficiency on a new graph corpus—is not derived from the paper's definitions; it is an empirical comparison in Section 5.4 (Figure 4, Table 1) against three baselines, including the same model without pre-training (Vassoyan et al.). The target corpus is not seen during pre-training, no target-task parameter is fitted into the pre-training objective, and the improvement is measured rather than assumed. The source and target tasks do share the same student simulator (Eqs. 1-4), which is a real threat to external validity but does not make the transfer result equivalent to its inputs by construction; the pre-trained model must still generalize across different documents, keywords, and prerequisite structures. The only self-citations are architectural and hyperparameter choices inherited from Vassoyan et al. (Section 3, Section 5.2, Appendix B); these are non-load-bearing for the transfer comparison, which uses the non-pre-trained version of the same architecture as a baseline. Appendix A.2's Eq. 6 does not in fact reduce to Eq. 3: for a known KC taught by an accessible document, Eq. 6 gives x'_i=1 with probability 1 while Eq. 3 gives 0, so the transition distribution can fail to normalize. This is an internal inconsistency in the core formalism and makes code inspection necessary, but it is a correctness/verifiability problem rather than a circular argument. Score 2 reflects only the minor non-load-bearing self-citation; no circular step is present.

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

The central claim rests on a simulated student environment whose dynamics are defined by the authors. The pre-training and fine-tuning tasks share this same dynamics, so the measured transfer gain may be partly an artifact of the simulator's internal consistency. No new physical entities are introduced.

free parameters (4)
  • E_pref edge probability p = 0.3
    Probability that a vertical dependency is added per student in the target corpus; chosen by hand in Section 5.1.
  • KC reward weights = row 1: 1, row 2: 2, row 3: 3
    Hand-assigned weights in the modified reward function (Section 5.2 and Appendix D.1) to favor difficult paths.
  • Decreasing exponential parameter = not specified
    Scenario 2 in Section 5.1 defines a decreasing exponential prior for KC count, but the rate parameter is not reported, making this a free choice.
  • Maximum episode length = 11
    Set to the number of major concepts (Section 5.2); bounding the horizon affects the return and the learning dynamics.
assumptions (5)
  • domain assumption A student's knowledge of each KC is binary, monotonic (never forgotten), and learning occurs iff all prerequisites are mastered (Eq. 2-3).
    Core of the simulated student dynamics; the entire evaluation rests on this model of learning.
  • domain assumption Student feedback is fully determined by prerequisite mastery and content mastery, giving exactly three outcomes (Eq. 1, 8).
    The observation function defines the reward signal; if real feedback is noisier, transfer results may differ.
  • domain assumption Sequential course corpora are well approximated by a chain of KCs, with each document teaching exactly one KC and no learning preferences (Section 2.3).
    This allows pre-training without expert annotation, but real sequential courses may not decompose so cleanly.
  • domain assumption Wikipedia2Vec embeddings and GPT-4 extracted keywords are sufficient semantic features for transfer (Section 3, Appendix F).
    The recommender only sees keyword nodes; if embeddings poorly represent the concepts, transfer would fail.
  • domain assumption The hand-designed grid corpus is representative of adaptive learning environments (Section 5.1, Appendix D.1).
    Results are reported on a single target corpus; a different structure could change the transfer benefit.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Pre-Trained Graph-Based Model for Adaptive Sequencing of Educational Documents." pith.science (2026). https://pith.science/paper/WQM52TON

@misc{pith2026241111520,
  author       = {Pith},
  title        = {Pith review of: A Pre-Trained Graph-Based Model for Adaptive Sequencing of Educational Documents},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WQM52TON}},
  note         = {Machine review of arXiv:2411.11520}
}
read the original abstract

Massive Open Online Courses (MOOCs) have greatly contributed to making education more accessible. However, many MOOCs maintain a rigid, one-size-fits-all structure that fails to address the diverse needs and backgrounds of individual learners. Learning path personalization aims to address this limitation, by tailoring sequences of educational content to optimize individual student learning outcomes. Existing approaches, however, often require either massive student interaction data or extensive expert annotation, limiting their broad application. In this study, we introduce a novel data-efficient framework for learning path personalization that operates without expert annotation. Our method employs a flexible recommender system pre-trained with reinforcement learning on a dataset of raw course materials. Through experiments on semi-synthetic data, we show that this pre-training stage substantially improves data-efficiency in a range of adaptive learning scenarios featuring new educational materials. This opens up new perspectives for the design of foundation models for adaptive learning.

Figures

Figures reproduced from arXiv: 2411.11520 by the authors.

Figure 1
Figure 1. Illustration of the partition of the domain: [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Illustration of the sampling process of a student-corpus pair. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Overview of the recommendation pipeline on one student [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Performance comparison (mean and confidence interval) depending on the pre-training [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: An intuitive view of student’s knowledge state model with keyword vectors [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: Learning curve and hyperparameters used for the RL pre-training stage [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: Illustration of the graph corpus we have used for the fine-tuning task. The nodes are the [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: Performance comparison of our model against two other pre-training strategies. [PITH_FULL_IMAGE:figures/full_fig_p016_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 27 canonical work pages

  1. [1]

    Gpt-4 technical report

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774, 2023

    arXiv 2023

  2. [2]

    Thompson sampling for contextual bandits with linear payoffs

    Shipra Agrawal and Navin Goyal. Thompson sampling for contextual bandits with linear payoffs. In International conference on machine learning, pages 127--135. PMLR, 2013

    work page 2013

  3. [3]

    The q-matrix method: Mining student response data for knowledge

    Tiffany Barnes. The q-matrix method: Mining student response data for knowledge. In American association for artificial intelligence 2005 educational data mining workshop, pages 1--8. AAAI Press, Pittsburgh, PA, USA, 2005

    work page 2005

  4. [4]

    Reinforcement learning for the adaptive scheduling of educational activities

    Jonathan Bassen, Bharathan Balaji, Michael Schaarschmidt, Candace Thille, Jay Painter, Dawn Zimmaro, Alex Games, Ethan Fast, and John C Mitchell. Reinforcement learning for the adaptive scheduling of educational activities. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems, pages 1--12, 2020

    work page 2020

  5. [5]

    The 2 sigma problem: The search for methods of group instruction as effective as one-to-one tutoring

    Benjamin S Bloom. The 2 sigma problem: The search for methods of group instruction as effective as one-to-one tutoring. Educational researcher, 13 0 (6): 0 4--16, 1984

    work page 1984

  6. [6]

    User models for adaptive hypermedia and adaptive educational systems

    Peter Brusilovsky and Eva Mill \'a n. User models for adaptive hypermedia and adaptive educational systems. In The adaptive web: methods and strategies of web personalization, pages 3--53. Springer, 2007

    work page 2007

  7. [7]

    LangChain , October 2022

    Harrison Chase. LangChain , October 2022. URL https://github.com/langchain-ai/langchain

    2022

  8. [8]

    Empirically evaluating the application of reinforcement learning to the induction of effective and adaptive pedagogical strategies

    Min Chi, Kurt VanLehn, Diane Litman, and Pamela Jordan. Empirically evaluating the application of reinforcement learning to the induction of effective and adaptive pedagogical strategies. User Modeling and User-Adapted Interaction, 21: 0 137--180, 2011

    work page 2011

Show all 43 references
  1. [9]

    Multi-armed bandits for intelligent tutoring systems

    Benjamin Clement, Didier Roy, Pierre-Yves Oudeyer, and Manuel Lopes. Multi-armed bandits for intelligent tutoring systems. Journal of Educational Data Mining, 7 0 (2), 2015

    2015

  2. [10]

    Knowledge tracing: Modeling the acquisition of procedural knowledge

    Albert T Corbett and John R Anderson. Knowledge tracing: Modeling the acquisition of procedural knowledge. User modeling and user-adapted interaction, 4: 0 253--278, 1994

    1994

  3. [11]

    Learning modular neural network policies for multi-task and multi-robot transfer

    Coline Devin, Abhishek Gupta, Trevor Darrell, Pieter Abbeel, and Sergey Levine. Learning modular neural network policies for multi-task and multi-robot transfer. In 2017 IEEE international conference on robotics and automation (ICRA), pages 2169--2176. IEEE, 2017

    2017

  4. [12]

    Spaces for the assessment of knowledge

    Jean-Paul Doignon and Jean-Claude Falmagne. Spaces for the assessment of knowledge. International journal of man-machine studies, 23 0 (2): 0 175--196, 1985

    1985

  5. [13]

    Matthias Fey and Jan E. Lenssen. Fast graph representation learning with PyTorch Geometric . In ICLR Workshop on Representation Learning on Graphs and Manifolds, 2019

    2019

  6. [14]

    Context-aware attentive knowledge tracing

    Aritra Ghosh, Neil Heffernan, and Andrew S Lan. Context-aware attentive knowledge tracing. In Proceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining, pages 2330--2339, 2020

    2020

  7. [15]

    Affective personalization of a social robot tutor for children’s second language skills

    Goren Gordon, Samuel Spaulding, Jacqueline Kory Westlund, Jin Joo Lee, Luke Plummer, Marayna Martinez, Madhurima Das, and Cynthia Breazeal. Affective personalization of a social robot tutor for children’s second language skills. In Proceedings of the AAAI conference on artific...

    2016

  8. [16]

    Soft actor-critic: Off-policy maximum entropy deep reinforcement learning with a stochastic actor

    Tuomas Haarnoja, Aurick Zhou, Pieter Abbeel, and Sergey Levine. Soft actor-critic: Off-policy maximum entropy deep reinforcement learning with a stochastic actor. In International conference on machine learning, pages 1861--1870. PMLR, 2018

    2018

  9. [17]

    Learning teaching strategies in an adaptive and intelligent educational system through reinforcement learning

    Ana Iglesias, Paloma Mart \' nez, Ricardo Aler, and Fernando Fern \'a ndez. Learning teaching strategies in an adaptive and intelligent educational system through reinforcement learning. Applied Intelligence, 31: 0 89--106, 2009

    2009

  10. [18]

    A contextual bandits framework for personalized learning action selection

    Andrew S Lan and Richard G Baraniuk. A contextual bandits framework for personalized learning action selection. In EDM, pages 424--429, 2016

    2016

  11. [19]

    Batch reinforcement learning

    Sascha Lange, Thomas Gabel, and Martin Riedmiller. Batch reinforcement learning. In Reinforcement learning: State-of-the-art, pages 45--73. Springer, 2012

    2012

  12. [20]

    The attribute hierarchy method for cognitive assessment: A variation on tatsuoka's rule-space approach

    Jacqueline P Leighton, Mark J Gierl, and Stephen M Hunka. The attribute hierarchy method for cognitive assessment: A variation on tatsuoka's rule-space approach. Journal of educational measurement, 41 0 (3): 0 205--237, 2004

    2004

  13. [21]

    Offline reinforcement learning: Tutorial, review, and perspectives on open problems

    Sergey Levine, Aviral Kumar, George Tucker, and Justin Fu. Offline reinforcement learning: Tutorial, review, and perspectives on open problems. arXiv preprint arXiv:2005.01643, 2020

    2005 arXiv

  14. [22]

    Playing atari with deep reinforcement learning

    Volodymyr Mnih, Koray Kavukcuoglu, David Silver, Alex Graves, Ioannis Antonoglou, Daan Wierstra, and Martin Riedmiller. Playing atari with deep reinforcement learning. arXiv preprint arXiv:1312.5602, 2013

    2013 arXiv

  15. [23]

    Asynchronous methods for deep reinforcement learning

    Volodymyr Mnih, Adria Puigdomenech Badia, Mehdi Mirza, Alex Graves, Timothy Lillicrap, Tim Harley, David Silver, and Koray Kavukcuoglu. Asynchronous methods for deep reinforcement learning. In International conference on machine learning, pages 1928--1937. PMLR, 2016

    1928

  16. [24]

    Model-based reinforcement learning: A survey

    Thomas M Moerland, Joost Broekens, Aske Plaat, Catholijn M Jonker, et al. Model-based reinforcement learning: A survey. Foundations and Trends in Machine Learning , 16 0 (1): 0 1--118, 2023

    2023

  17. [25]

    A model-free affective reinforcement learning approach to personalization of an autonomous social robot companion for early literacy education

    Hae Won Park, Ishaan Grover, Samuel Spaulding, Louis Gomez, and Cynthia Breazeal. A model-free affective reinforcement learning approach to personalization of an autonomous social robot companion for early literacy education. In Proceedings of the AAAI Conference on Artificial...

    2019

  18. [26]

    Deep knowledge tracing

    Chris Piech, Jonathan Bassen, Jonathan Huang, Surya Ganguli, Mehran Sahami, Leonidas J Guibas, and Jascha Sohl-Dickstein. Deep knowledge tracing. In C. Cortes, N. Lawrence, D. Lee, M. Sugiyama, and R. Garnett, editors, Advances in Neural Information Processing Systems, volume ...

    2015

  19. [27]

    Faster teaching via pomdp planning

    Anna N Rafferty, Emma Brunskill, Thomas L Griffiths, and Patrick Shafto. Faster teaching via pomdp planning. Cognitive science, 40 0 (6): 0 1290--1332, 2016

    2016

  20. [28]

    Accelerating human learning with deep reinforcement learning

    Siddharth Reddy, Sergey Levine, and Anca Dragan. Accelerating human learning with deep reinforcement learning. In NIPS workshop: teaching machines, robots, and humans, 2017

    2017

  21. [29]

    Progressive neural networks

    Andrei A Rusu, Neil C Rabinowitz, Guillaume Desjardins, Hubert Soyer, James Kirkpatrick, Koray Kavukcuoglu, Razvan Pascanu, and Raia Hadsell. Progressive neural networks. NIPS Deep Learning Symposium, 2016

    2016

  22. [30]

    Proximal policy optimization algorithms

    John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347, 2017

    2017 arXiv

  23. [31]

    Curriculumtutor: An adaptive algorithm for mastering a curriculum

    KM Shabana, Chandrashekar Lakshminarayanan, and Jude K Anil. Curriculumtutor: An adaptive algorithm for mastering a curriculum. In International Conference on Artificial Intelligence in Education, pages 319--331. Springer, 2022

    2022

  24. [32]

    Masked label prediction: Unified message passing model for semi-supervised classification

    Yunsheng Shi, Zhengjie Huang, Shikun Feng, Hui Zhong, Wenjing Wang, and Yu Sun. Masked label prediction: Unified message passing model for semi-supervised classification. In Zhi-Hua Zhou, editor, Proceedings of the Thirtieth International Joint Conference on Artificial Intelli...

    2021 doi

  25. [33]

    Deep reinforcement learning to simulate, train, and evaluate instructional sequencing policies

    Jithendaraa Subramanian and Jack Mostow. Deep reinforcement learning to simulate, train, and evaluate instructional sequencing policies. In Spotlight presentation at Reinforcement Learning for Education workshop at Educational Data Mining 2021 conference, 2021

    2021

  26. [34]

    Policy gradient methods for reinforcement learning with function approximation

    Richard S Sutton, David McAllester, Satinder Singh, and Yishay Mansour. Policy gradient methods for reinforcement learning with function approximation. Advances in neural information processing systems, 12, 1999

    1999

  27. [35]

    Deepmind control suite

    Yuval Tassa, Yotam Doron, Alistair Muldal, Tom Erez, Yazhe Li, Diego de Las Casas, David Budden, Abbas Abdolmaleki, Josh Merel, Andrew Lefrancq, et al. Deepmind control suite. arXiv preprint arXiv:1801.00690, 2018

    2018 arXiv

  28. [36]

    Recommending remedial readings using student knowledge state

    Khushboo Thaker, Lei Zhang, Daqing He, and Peter Brusilovsky. Recommending remedial readings using student knowledge state. International Educational Data Mining Society, 2020

    2020

  29. [37]

    Terry, Ariel Kwiatkowski, John U

    Mark Towers, Jordan K. Terry, Ariel Kwiatkowski, John U. Balis, Gianluca de Cola, Tristan Deleu, Manuel Goulão, Andreas Kallinteris, Arjun KG, Markus Krimmel, Rodrigo Perez-Vicente, Andrea Pierré, Sander Schulhoff, Jun Jet Tai, Andrew Tan Jin Shen, and Omar G. Younis. Gymnasiu...

    2023

  30. [38]

    Towards Scalable Adaptive Learning with Graph Neural Networks and Reinforcement Learning

    Jean Vassoyan, Jill-Jênn Vie, and Pirmin Lemberger. Towards Scalable Adaptive Learning with Graph Neural Networks and Reinforcement Learning . In Proceedings of the 16th International Conference on Educational Data Mining , pages 351--361. International Educational Data Mining...

    2023 doi

  31. [39]

    Neural graph collaborative filtering

    Xiang Wang, Xiangnan He, Meng Wang, Fuli Feng, and Tat-Seng Chua. Neural graph collaborative filtering. In Proceedings of the 42nd international ACM SIGIR conference on Research and development in Information Retrieval, pages 165--174, 2019

    2019

  32. [40]

    Tianshou: A highly modularized deep reinforcement learning library

    Jiayi Weng, Huayu Chen, Dong Yan, Kaichao You, Alexis Duburcq, Minghao Zhang, Yi Su, Hang Su, and Jun Zhu. Tianshou: A highly modularized deep reinforcement learning library. Journal of Machine Learning Research, 23 0 (267): 0 1--6, 2022. URL http://jmlr.org/papers/v23/21-1127.html

    2022

  33. [41]

    W ikipedia2 V ec: An efficient toolkit for learning and visualizing the embeddings of words and entities from W ikipedia

    Ikuya Yamada, Akari Asai, Jin Sakuma, Hiroyuki Shindo, Hideaki Takeda, Yoshiyasu Takefuji, and Yuji Matsumoto. W ikipedia2 V ec: An efficient toolkit for learning and visualizing the embeddings of words and entities from W ikipedia. In Proceedings of the 2020 Conference on Emp...

    2020 doi

  34. [42]

    Decoupling dynamics and reward for transfer learning

    Amy Zhang, Harsh Satija, and Joelle Pineau. Decoupling dynamics and reward for transfer learning. In 6th International Conference on Learning Representations, ICLR 2018, Vancouver, BC, Canada, April 30 - May 3, 2018, Workshop Track Proceedings . OpenReview.net, 2018. URL https...

    2018

  35. [43]

    Transfer learning in deep reinforcement learning: A survey

    Zhuangdi Zhu, Kaixiang Lin, Anil K Jain, and Jiayu Zhou. Transfer learning in deep reinforcement learning: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023

    2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.