Pith. sign in

REVIEW 4 major objections 6 minor 3 cited by

Graph Domain Adaptation with Dual-branch Encoder and Two-level Alignment for Whole Slide Image-based Survival Prediction

T0 review · 4 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 DETA, a dual-branch graph encoder with category and feature alignment, transfers WSI survival models across cancer domains and reports the best C-index on all tested TCGA directions.

desk verdict First GDA treatment of WSI survival with solid experiments, but a shaky category-alignment derivation and no outside-TCGA validation keep it from being trustworthy as-is. read the letter →

arxiv 2411.14001 v1 pith:YC5DBBTC submitted 2024-11-21 cs.CV

Yuntao Shou , Peiqiang Yan , Xingjian Yuan , Xiangyong Cao , Qian Zhao , Deyu Meng This is my paper · ORCID
classification cs.CV
keywords graphdomainadaptationwholeslideimagesurvivalpredictiondual-branchencodercategoryalignmentfeatureadversarialperturbationTCGA
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 argues that the domain shift between whole slide image datasets from different hospitals can be modeled as a graph domain adaptation problem, and that aligning both category and feature distributions lets a survival model trained on labeled source slides transfer to an unlabeled target cohort. The proposed DETA framework pairs a dual-branch graph encoder (message passing plus shortest path) with two-level alignment: coupled pseudo-label supervision between branches for category alignment, and adversarial perturbations on source features for feature alignment. On four TCGA cancer datasets, the paper reports the best cross-domain C-index in every adaptation direction tested, with gains over prior WSI survival baselines. If this holds, clinical sites could reuse models trained elsewhere without collecting survival labels in the target domain.

What carries the argument

The load-bearing machinery is the dual-branch graph encoder together with the branch-coupling ELBO objective. The message passing (MP) branch follows GCN-style neighborhood aggregation, while the shortest path (SP) branch aggregates features over nodes reachable by shortest paths of length up to K with sinusoidal position encodings; the two branches parameterize two distributions $p_\theta$ and $q_\phi$ over risk labels. Coupling means each branch's high-confidence pseudo-labels on the target domain train the other branch's predictor, and the two ELBO losses $L_1$ and $L_2$ are optimized alternately, which the paper claims reduces the divergence between source and target category distributions. The feature-level piece is an adversarial perturbation: bounded perturbations $\delta_{MP}$ and $\delta_{SP}$ added to source graph node features are learned against a domain classifier $D$ so that source features are pushed toward target feature statistics.

What would settle it

Train DETA on a source-target pair while shuffling the pseudo-labels within each confidence-filtered batch, preserving the same class frequencies; if the C-index does not drop materially, the category-coupling term is not doing the claimed work. Alternatively, a variant with feature alignment only should be clearly worse than full DETA; if it matches DETA, the category level is redundant.

Watch

Extended reading notes

Core claim

The central claim is that WSI survival prediction across domains is improved by explicitly aligning both the risk-level category distributions and the feature distributions of source and target graphs. DETA trains two graph encoders on the same WSI graphs—a message passing branch that implicitly aggregates neighborhood semantics and a shortest path branch that explicitly aggregates high-order path structure—and uses the dual structure to derive two coupled evidence lower bounds: each branch supervises the other's target pseudo-labels after confidence filtering. Simultaneously, learnable perturbations bounded by epsilon are added to source node features and optimized adversarially against a domain classifier, so the source feature distribution is pushed toward the target. The paper reports that DETA outperforms all compared WSI survival methods on all twelve source-to-target directions across BLCA, LGG, LUAD, and UCEC datasets.

Load-bearing premise

The method assumes that high-confidence pseudo-labels produced by one branch on the target domain are accurate enough to teach the other branch; if the domain shift biases those labels, both branches can converge on the same wrong categories, and the reported alignment would be self-confirmation rather than correction.

Editorial extensions

If this is right

  • Unlabeled target hospitals could be adapted to using DETA without collecting survival labels on site.
  • The dual-branch coupling gives a template for category alignment in other graph-based medical tasks where target labels are absent.
  • Feature-level adversarial perturbation becomes a reusable augmentation strategy for cross-domain graph learning beyond survival analysis.
  • Because the formulation uses discrete risk levels, the category alignment directly targets the clinical stratification quantity used in practice.
  • The reported C-index margins over strong baselines suggest that alignment, not just the encoder architecture, drives the transfer gains.

Reading between the lines

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

  • The paper does not isolate the coupling objective from feature alignment with a single ablation; a reader would need a 'feature alignment only' variant to know how much the category coupling contributes, and this reader predicts that contribution grows when target pseudo-labels are accurate.
  • The shortest path branch's explicit high-order aggregation may matter most when tissue topology, not just patch appearance, shifts across sites; a testable prediction is that DETA's advantage widens as graph structure divergence increases.
  • Since the pseudo-label confidence threshold $\zeta$ is a free parameter, deployment would need a no-target-label rule for setting it; a source-only validation split could serve that role.
  • If the coupling locks both branches onto the same biased pseudo-labels, the reported category alignment could be self-generated rather than truthful; measuring agreement between each branch and true target labels on a held-out annotated set would separate these cases.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes DETA, a graph domain adaptation framework for whole-slide-image-based survival prediction. DETA uses a dual-branch graph encoder (message-passing and shortest-path branches) and performs two levels of alignment: category-level alignment via coupled evidence lower bounds and confidence-filtered pseudo-labels, and feature-level alignment via adversarial source-feature perturbations. Experiments on four TCGA datasets in a one-source-three-target protocol report consistent improvements in C-index over existing MIL and survival-analysis baselines, with ablations attributing gains to the dual-branch design, the perturbation modules, and the branch-coupling objective.

Significance. If the empirical results and the claimed mechanism are sound, DETA would be a useful contribution to cross-domain WSI survival analysis, a realistic setting where target hospitals may lack survival annotations. The paper introduces a nontrivial dual-branch coupling idea and an adversarial perturbation strategy, and it includes extensive experiments, ablations, and visualizations. However, the central theoretical derivation in Section 4.3 is not valid as written, the pseudo-label-based alignment is potentially circular and is not validated against true target labels, and the experimental protocol is underspecified regarding whether baselines also receive target unlabeled data. Because these points bear directly on whether the reported gains reflect the proposed alignment mechanisms, the contribution is not yet conclusively established.

major comments (4)
  1. [§4.3, Eq. (10)] Equation (10) is presented as an equivalent rewriting of the ELBO in Eq. (9), but the identity is not derived and appears dimensionally inconsistent: the left-hand side is a KL divergence between q(y^t|G^t) and p(y^t|G^s,G^t,y^s), while the right-hand side is an expectation of log p(y^t|G^t) plus a source-domain expectation over q. No standard ELBO identity yields this form, and the citation to [73] does not explain the step. Because the subsequent losses L1 and L2 in Eq. (11) are asserted to follow from this relation, the theoretical grounding of the category-level alignment is unsupported. The authors should either provide a correct derivation of L1/L2 from a valid ELBO, or present them as a heuristic pseudo-label consistency loss and validate the choice empirically.
  2. [§4.3, Eq. (11), Figs. 5-6] The category-level alignment relies on confidence-filtered pseudo-labels produced by the model itself, and the alignment plots in Figs. 5-6 compare those self-generated distributions. Since both branches are initialized from the same source-trained encoder and receive no target supervision, their target predictions are correlated; the confidence filter (threshold ζ) preferentially retains samples on which the model is overconfident, which under domain shift can be systematically biased. The paper does not report any measure of pseudo-label accuracy on the target domain or an oracle experiment with true target labels. Please add (i) target pseudo-label accuracy versus confidence threshold, and (ii) an upper-bound experiment in which target labels are used for the branch-coupling loss, to demonstrate that the coupling aligns true risk categories rather than merely making the two branches agree with each other.
  3. [§5.2, Table 1] The comparison in Table 1 is not fully specified. DETA is fine-tuned on unlabeled target data during the GDA stage, but the text does not state whether the baseline methods (AttMIL, CLAM, TransMIL, etc.) also receive target-domain data in any form. If the baselines are trained only on the source domain and evaluated on target, the reported gains may reflect access to target unlabeled data rather than the proposed alignment mechanism. Please clarify the protocol for each baseline (source-only or UDA-adapted), and if they are source-only, include a source-only variant of DETA and/or UDA-adapted baselines for a fair comparison.
  4. [§4.4, Eq. (12), §4.5, Eq. (13)] The training objective in Eq. (13) minimizes over δ_MP, δ_SP, h, and an unexplained quantity S, but it does not explicitly include the graph encoder F(·), despite the text in §4.5 stating that both F(·) and h(·) are fine-tuned. In Eq. (12), the perturbation δ is optimized in a minimax fashion with the domain classifier D, but it is unclear whether minimizing L_AP with respect to δ encourages source features to become more target-like or merely changes the source features without updating the encoder. Please specify the exact optimization variables in each stage, including the update of F, and define S. Also align the notation for the domain classifier, which is T(·) in §4.1 and Fig. 2 but D in Eq. (12).
minor comments (6)
  1. [§4.1] The section heading contains a typo: 'Fraemwork Overview' should be 'Framework Overview'.
  2. [Fig. 1] In the caption of Fig. 1, the label 'BLCALUAD' appears to be a corrupted concatenation of the two dataset names.
  3. [§4.1 and §4.4] The domain classifier is introduced as T(·) in §4.1 and Fig. 2 but is denoted D in Eq. (12); use a consistent symbol throughout.
  4. [§5.1] The paper does not report the values of the hyperparameters K, ε, ζ, the number of discrete risk levels, or the KNN neighborhood size used in the experiments; these should be listed for reproducibility.
  5. [Abstract and Introduction] The claim of being 'the first attempt to alleviate the domain shift issue for WSI data analysis' is too broad, as there is substantial prior work on stain normalization and domain adaptation in histopathology; please narrow the claim to the specific task of WSI-based survival analysis under unsupervised graph domain adaptation and cite the closest prior work.
  6. [§3, Eq. (5)] The survival likelihood in Eq. (5) should be clarified: for uncensored observations the standard contribution is log h(y_i|x_i) + log S(y_i|x_i), and for censored observations log S(y_i+1|x_i); the current expression is correct only if the time intervals and censoring notation are precisely defined.

Circularity Check

1 steps flagged · score 5.0 of 10

Category-level alignment in Eqs. (9)-(11) is built on the model's own confidence-filtered pseudo-labels, so the claimed alignment of target category distributions is partly self-referential; the final C-index benchmark against held-out target labels provides an external check.

  1. self definitional [Section 4.3, Eq. (11) (branch-coupling losses)]
    "Then denoting the distributions parameterized by MP and SP as pθ and qϕ, using the coupling strategy introduced before, we get the following losses for updating MP and SP, respectively, with the other fixed: L1 = − Epθ(ˆyt_i |Gs,Gt,ys)>ζ [log qϕ(ˆyt_i | Gt_i)] − Epθ(ys,Gs) log pθ(ys_i | Gs_i); L2 = − Eqϕ(ˆyt_i |Gs,Gt,ys)>ζ [log pθ(ˆyt_i | Gt_i)] − Eqϕ(ys,Gs) log qϕ(ys_i | Gs_i) where ˆyt_i is the target graph pseudo-label filtered by the MP or SP branch."

    The target pseudo-label ŷt_i used as the training target is drawn from pθ or qϕ, i.e., from the very branches being trained. The category-level alignment therefore minimizes divergence between one branch's predictions and the other branch's confidence-filtered predictions, not between the source and true target category distributions. The paper itself states that pseudo-labels 'could be inaccurate and have a category distribution different from that in the source domain,' yet the coupling objective aligns two self-generated distributions. Because both branches are pretrained on the same source survival loss, their errors are correlated, so confidence filtering preferentially retains samples where the model is overconfident under domain shift.

full rationale

The paper's headline empirical claim is evaluated on four TCGA datasets with C-index computed against held-out target ground-truth labels, so the central survival-prediction result does not reduce to the model's own outputs. The one structural weakness is the category-level alignment: Eq. (11) trains each branch to predict the other branch's confidence-filtered pseudo-labels, making the aligned target category distribution self-generated rather than anchored to true target categories. The ELBO 'equivalence' in Eq. (10), attributed to [73], is also dimensionally suspect and is not a valid identity as written, which undermines the claimed first-principles derivation; that is a correctness concern rather than a circularity. Because the final evaluation is external, this is partial, not full, circularity.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The method relies on several unreported hyperparameters and on the validity of a coupled ELBO objective that is not fully derived. No new physical or biological entities are introduced.

free parameters (5)
  • K (shortest path length) = not reported
    Maximum path length for the SP branch in Section 4.2; no value or tuning procedure given.
  • epsilon (perturbation bound) = not reported
    Norm bound for delta_MP and delta_SP in Eq (12); no value given.
  • zeta (pseudo-label confidence threshold) = not reported
    Threshold for filtering pseudo-labels in Eq (11); no value given.
  • number of discrete risk levels = not reported
    Discretization of survival time into K risk levels in Eq (2); K is not specified and is overloaded with the path-length K.
  • KNN neighborhood size for graph construction = not reported
    Number of neighbors in KNN graph construction after CLAM feature extraction (Section 3); no value given.
assumptions (6)
  • domain assumption Unlabeled target-domain graphs are available during training and can be used to fine-tune the shared encoder and predictor.
    The GDA formulation in Section 3 requires target data D^t without labels; the method uses them for pseudo-labeling and adversarial alignment.
  • domain assumption Source and target domains share the same node feature space and the same discrete risk-level label space.
    A single dual-branch encoder and hazard predictor are shared across domains in Section 4.1.
  • ad hoc to paper The two graph branches can be used as mutual variational distributions p and q, and exchanging their roles yields the coupling objectives L1 and L2.
    Section 4.3 asserts the coupling from the ELBO in Eq (9) to Eq (11) without a complete derivation.
  • domain assumption Confidence-filtered pseudo-labels from one branch are sufficiently accurate to train the other branch.
    The filtering threshold zeta in Eq (11) is meant to select high-quality pseudo-labels, but no accuracy validation is provided.
  • ad hoc to paper The shortest path branch with positional encoding captures explicit high-order semantic structure that improves cross-domain alignment.
    Section 4.2 introduces the SP branch as a design choice; its benefit is only shown empirically in ablations.
  • domain assumption The adversarial perturbation min-max game in Eq (12) aligns feature distributions.
    This is a standard adversarial domain adaptation assumption, here applied to source feature perturbations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Graph Domain Adaptation with Dual-branch Encoder and Two-level Alignment for Whole Slide Image-based Survival Prediction." pith.science (2026). https://pith.science/paper/YC5DBBTC

@misc{pith2026241114001,
  author       = {Pith},
  title        = {Pith review of: Graph Domain Adaptation with Dual-branch Encoder and Two-level Alignment for Whole Slide Image-based Survival Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YC5DBBTC}},
  note         = {Machine review of arXiv:2411.14001}
}
read the original abstract

In recent years, histopathological whole slide image (WSI)- based survival analysis has attracted much attention in medical image analysis. In practice, WSIs usually come from different hospitals or laboratories, which can be seen as different domains, and thus may have significant differences in imaging equipment, processing procedures, and sample sources. These differences generally result in large gaps in distribution between different WSI domains, and thus the survival analysis models trained on one domain may fail to transfer to another. To address this issue, we propose a Dual-branch Encoder and Two-level Alignment (DETA) framework to explore both feature and category-level alignment between different WSI domains. Specifically, we first formulate the concerned problem as graph domain adaptation (GDA) by virtue the graph representation of WSIs. Then we construct a dual-branch graph encoder, including the message passing branch and the shortest path branch, to explicitly and implicitly extract semantic information from the graph-represented WSIs. To realize GDA, we propose a two-level alignment approach: at the category level, we develop a coupling technique by virtue of the dual-branch structure, leading to reduced divergence between the category distributions of the two domains; at the feature level, we introduce an adversarial perturbation strategy to better augment source domain feature, resulting in improved alignment in feature distribution. To the best of our knowledge, our work is the first attempt to alleviate the domain shift issue for WSI data analysis. Extensive experiments on four TCGA datasets have validated the effectiveness of our proposed DETA framework and demonstrated its superior performance in WSI-based survival analysis.

Figures

Figures reproduced from arXiv: 2411.14001 by the authors.

Figure 1
Figure 1. An example of pathology images with distribution shift. The left shows image from two WSI datasets, i.e., BLCA and LUAD. The right shows the category distribution and the t-SNE visualization of the extracted features by CLAM [30] method. 1. Introduction In clinic, survival analysis aims to predict the time from the start of treatment to death or disease recurrence, which plays an important role in clinical prognosis… view at source ↗
Figure 2
Figure 2. An overview of our proposed archicture DETA. The dual branch graph encoder utilizes the MP branch and the SP branch to [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. An overview of the proposed branch coupling. We ad [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: According to the predicted risks, all patients are stratified into low-risk and high-risk group, and Kaplan-Meier analysis is used [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 6
Figure 6. Figure 6: Visualizations of the category distributions of source and [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

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

  1. SDR-GNN: Spectral Domain Reconstruction Graph Neural Network for Incomplete Multimodal Learning in Conversational Emotion Recognition

    cs.CL 2024-11 reject novelty 5.0 of 10

    SDR-GNN is a graph neural network that reconstructs missing multimodal features and labels utterance emotions, with reported gains over prior methods that are inconsistent across datasets.

  2. GroupFace: Imbalanced Age Estimation Based on Multi-hop Attention Graph Convolutional Network and Group-aware Margin Optimization

    cs.CV 2024-12 reject novelty 4.0 of 10

    GroupFace combines a multi-hop attention graph network with a reinforcement-learning margin scheduler for imbalanced face age estimation, reporting modest benchmark gains but with internal inconsistencies in the rewar...

  3. Dynamic Graph Neural ODE Network for Multi-modal Emotion Recognition in Conversation

    cs.CL 2024-12 reject novelty 4.0 of 10

    DGODE combines adaptive mixhop aggregation with a graph ODE for multimodal emotion recognition in conversation, reporting SOTA numbers on IEMOCAP and MELD, but the supporting derivation and experimental reporting are ...

Reference graph

Works this paper leans on

82 extracted references · 64 canonical work pages · cited by 3 Pith papers

  1. [73]

    Dream: Dual structured ex- ploration with mixup for open-set graph domain adaption

    Nan Yin, Mengzhu Wang, Zhenghan Chen, Li Shen, Huan Xiong, Bin Gu, and Xiao Luo. Dream: Dual structured ex- ploration with mixup for open-set graph domain adaption. In The Twelfth International Conference on Learning Repre- sentations, 2024. 5

    work page 2024

  2. [1]

    Pathalign: A vision–language model for whole slide images in histopathology

    Faruk Ahmed, Andrew Sellergren, Lin Yang, Jinhua Xu, Boris Babenko, Abbi Ward, Niels Olson, Arash Mo- htashamian, Yossi Matias, Greg Corrado, et al. Pathalign: A vision–language model for whole slide images in histopathology. In MICCAI Workshop on Computational Pathology with Multimodal Data (COMPAYL). 2, 6

  3. [2]

    Der-gcn: Dialogue and event relation-aware graph convo- lutional neural network for multimodal dialogue emotion recognition

    Wei Ai, Yuntao Shou, Tao Meng, Nan Yin, and Keqin Li. Der-gcn: Dialogue and event relation-aware graph convo- lutional neural network for multimodal dialogue emotion recognition. arXiv preprint arXiv:2312.10579, 2023. 2

    arXiv 2023

  4. [3]

    A two-stage multimodal emotion recog- nition model based on graph contrastive learning

    Wei Ai, FuChen Zhang, Tao Meng, YunTao Shou, HongEn Shao, and Keqin Li. A two-stage multimodal emotion recog- nition model based on graph contrastive learning. In 2023 IEEE 29th International Conference on Parallel and Dis- tributed Systems (ICPADS), pages 397–404. IEEE, 2023. 3

    2023

  5. [4]

    Mcsff: Multi-modal consistency and speci- ficity fusion framework for entity alignment

    Wei Ai, Wen Deng, Hongyi Chen, Jiayi Du, Tao Meng, and Yuntao Shou. Mcsff: Multi-modal consistency and speci- ficity fusion framework for entity alignment. arXiv preprint arXiv:2410.14584, 2024. 3

    arXiv 2024

  6. [5]

    Graph contrastive learning via cluster- refined negative sampling for semi-supervised text classifi- cation

    Wei Ai, Jianbin Li, Ze Wang, Jiayi Du, Tao Meng, Yuntao Shou, and Keqin Li. Graph contrastive learning via cluster- refined negative sampling for semi-supervised text classifi- cation. arXiv preprint arXiv:2410.18130, 2024. 3

    arXiv 2024

  7. [6]

    Der-gcn: Dialog and event relation-aware graph convolutional neural network for multimodal dialog emotion recognition

    Wei Ai, Yuntao Shou, Tao Meng, and Keqin Li. Der-gcn: Dialog and event relation-aware graph convolutional neural network for multimodal dialog emotion recognition. IEEE Transactions on Neural Networks and Learning Systems ,

    work page

  8. [7]

    Edge-enhanced minimum-margin graph attention network for short text classification

    Wei Ai, Yingying Wei, Hongen Shao, Yuntao Shou, Tao Meng, and Keqin Li. Edge-enhanced minimum-margin graph attention network for short text classification. Expert Systems with Applications, 251:124069, 2024. 3

    work page 2024

Show all 82 references
  1. [8]

    Long-term follow-up of pediatric ms patients starting treat- ment with injectable first-line agents: a multicentre, italian, retrospective, observational study

    Damiano Baroncini, Mauro Zaffaroni, Lucia Moiola, Lorena Lorefice, Giuseppe Fenu, Pietro Iaffaldano, Marta Simone, Fulvia Fanelli, Francesco Patti, Emanuele D’Amico, et al. Long-term follow-up of pediatric ms patients starting treat- ment with injectable first-line agents: a m...

    2019

  2. [9]

    Shortest-path kernels on graphs

    Karsten M Borgwardt and Hans-Peter Kriegel. Shortest-path kernels on graphs. In Fifth IEEE International Conference on Data Mining (ICDM’05), pages 8–pp. IEEE, 2005. 4, 5

    2005

  3. [10]

    Assessing long- term prognosis improvement as a consequence of treatment pattern changes in ms

    Ruggero Capra, Cinzia Cordioli, Sarah Rasia, Fabio Gallo, Alessio Signori, and Maria Pia Sormani. Assessing long- term prognosis improvement as a consequence of treatment pattern changes in ms. Multiple Sclerosis Journal, 23(13): 1757–1761, 2017. 1, 2

    2017

  4. [11]

    Scaling vision transformers to gigapixel images via hierarchical self-supervised learning

    Richard J Chen, Chengkuan Chen, Yicong Li, Tiffany Y Chen, Andrew D Trister, Rahul G Krishnan, and Faisal Mahmood. Scaling vision transformers to gigapixel images via hierarchical self-supervised learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Patter...

    2022

  5. [12]

    Towards a general-purpose foundation model for computational pathology

    Richard J Chen, Tong Ding, Ming Y Lu, Drew FK Williamson, Guillaume Jaume, Andrew H Song, Bowen Chen, Andrew Zhang, Daniel Shao, Muhammad Shaban, et al. Towards a general-purpose foundation model for computational pathology. Nature Medicine, 30(3):850–862,

  6. [13]

    Graph transfer learning via adversarial domain adaptation with graph convolution

    Quanyu Dai, Xiao-Ming Wu, Jiaren Xiao, Xiao Shen, and Dan Wang. Graph transfer learning via adversarial domain adaptation with graph convolution. IEEE Transactions on Knowledge and Data Engineering , 35(5):4908–4922, 2022. 3

    2022

  7. [14]

    Generating hypergraph-based high-order representations of whole-slide histopathological images for survival prediction

    Donglin Di, Changqing Zou, Yifan Feng, Haiyan Zhou, Rongrong Ji, Qionghai Dai, and Yue Gao. Generating hypergraph-based high-order representations of whole-slide histopathological images for survival prediction. IEEE Transactions on Pattern Analysis and Machine Intelligence, 4...

    2022

  8. [15]

    Pathology-and-genomics multimodal transformer for survival outcome prediction

    Kexin Ding, Mu Zhou, Dimitris N Metaxas, and Shaoting Zhang. Pathology-and-genomics multimodal transformer for survival outcome prediction. In International Conference on Medical Image Computing and Computer-Assisted Interven- tion, pages 622–631. Springer, 2023. 7

    2023

  9. [16]

    A closer look at distribution shifts and out- of-distribution generalization on graphs

    Mucong Ding, Kezhi Kong, Jiuhai Chen, John Kirchen- bauer, Micah Goldblum, David Wipf, Furong Huang, and Tom Goldstein. A closer look at distribution shifts and out- of-distribution generalization on graphs. In NeurIPS 2021 Workshop on Distribution Shifts: Connecting Methods a...

    2021

  10. [17]

    Towards open temporal graph neural networks

    Kaituo Feng, Changsheng Li, Xiaolu Zhang, and JUN ZHOU. Towards open temporal graph neural networks. In The Eleventh International Conference on Learning Repre- sentations, 2023. 3

    2023

  11. [18]

    Chest ct score in covid- 19 patients: correlation with disease severity and short-term prognosis

    Marco Francone, Franco Iafrate, Giorgio Maria Masci, Si- mona Coco, Francesco Cilia, Lucia Manganaro, Valeria Panebianco, Chiara Andreoli, Maria Chiara Colaiacomo, Maria Antonella Zingaropoli, et al. Chest ct score in covid- 19 patients: correlation with disease severity and s...

    2020

  12. [19]

    Topology-aware graph pooling networks

    Hongyang Gao, Yi Liu, and Shuiwang Ji. Topology-aware graph pooling networks. IEEE Transactions on Pattern Anal- ysis and Machine Intelligence, 43(12):4512–4518, 2021. 4, 5

    2021

  13. [20]

    Learning adaptive node embeddings across graphs

    Gaoyang Guo, Chaokun Wang, Bencheng Yan, Yunkai Lou, Hao Feng, Junchao Zhu, Jun Chen, Fei He, and S Yu Philip. Learning adaptive node embeddings across graphs. IEEE Transactions on Knowledge and Data Engineering , 35(6): 6028–6042, 2022. 3 9

    2022

  14. [21]

    Attention-based deep multiple instance learning

    Maximilian Ilse, Jakub Tomczak, and Max Welling. Attention-based deep multiple instance learning. In Inter- national conference on machine learning, pages 2127–2136. PMLR, 2018. 7

    2018

  15. [22]

    Assessing changes in relapse rates in multiple sclerosis.Multiple sclerosis journal, 16(12): 1414–1421, 2010

    Seidu Inusah, Maria P Sormani, Stacey S Cofield, Inmac- ulada B Aban, Solomon K Musani, Vinodh Srinivasasaina- gendra, and Gary R Cutter. Assessing changes in relapse rates in multiple sclerosis.Multiple sclerosis journal, 16(12): 1414–1421, 2010. 1

    2010

  16. [23]

    Transcriptomics-guided slide representation learning in computational pathology

    Guillaume Jaume, Lukas Oldenburg, Anurag Vaidya, Richard J Chen, Drew FK Williamson, Thomas Peeters, An- drew H Song, and Faisal Mahmood. Transcriptomics-guided slide representation learning in computational pathology. In Proceedings of the IEEE/CVF Conference on Computer Vi- ...

    2024

  17. [24]

    Model- ing dense multimodal interactions between biological path- ways and histology for survival prediction

    Guillaume Jaume, Anurag Vaidya, Richard J Chen, Drew FK Williamson, Paul Pu Liang, and Faisal Mahmood. Model- ing dense multimodal interactions between biological path- ways and histology for survival prediction. InProceedings of the IEEE/CVF Conference on Computer Vision and ...

    2024

  18. [25]

    Machine learning and deep learning ap- proaches in breast cancer survival prediction using clinical data

    EY Kalafi, NAM Nor, NA Taib, MD Ganggayah, C Town, and SK Dhillon. Machine learning and deep learning ap- proaches in breast cancer survival prediction using clinical data. Folia biologica, 65(5-6):212–220, 2019. 1, 2

    2019

  19. [26]

    Dual-stream multiple instance learning network for whole slide image classifica- tion with self-supervised contrastive learning

    Bin Li, Yin Li, and Kevin W Eliceiri. Dual-stream multiple instance learning network for whole slide image classifica- tion with self-supervised contrastive learning. In Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14318–14328, 2021. 7

    2021

  20. [27]

    Dynamic graph repre- sentation with knowledge-aware attention for histopathology whole slide image analysis

    Jiawen Li, Yuxuan Chen, Hongbo Chu, Qiehe Sun, Tian Guan, Anjia Han, and Yonghong He. Dynamic graph repre- sentation with knowledge-aware attention for histopathology whole slide image analysis. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognitio...

    2024

  21. [28]

    Graph cnn for survival analysis on whole slide pathological images

    Ruoyu Li, Jiawen Yao, Xinliang Zhu, Yeqing Li, and Jun- zhou Huang. Graph cnn for survival analysis on whole slide pathological images. In International Conference on Med- ical Image Computing and Computer-Assisted Intervention , pages 174–182. Springer, 2018. 2, 3

    2018

  22. [29]

    Multi-domain generalized graph meta learning

    Mingkai Lin, Wenzhong Li, Ding Li, Yizhou Chen, Guohao Li, and Sanglu Lu. Multi-domain generalized graph meta learning. In Proceedings of the AAAI Conference on Artifi- cial Intelligence, pages 4479–4487, 2023. 3

    2023

  23. [30]

    Data-efficient and weakly supervised computational pathology on whole- slide images

    Ming Y Lu, Drew FK Williamson, Tiffany Y Chen, Richard J Chen, Matteo Barbieri, and Faisal Mahmood. Data-efficient and weakly supervised computational pathology on whole- slide images. Nature biomedical engineering, 5(6):555–570,

  24. [31]

    Visual language pre- trained multiple instance zero-shot transfer for histopathol- ogy images

    Ming Y Lu, Bowen Chen, Andrew Zhang, Drew FK Williamson, Richard J Chen, Tong Ding, Long Phi Le, Yung- Sung Chuang, and Faisal Mahmood. Visual language pre- trained multiple instance zero-shot transfer for histopathol- ogy images. In Proceedings of the IEEE/CVF conference on C...

    2023

  25. [32]

    Source-free progressive graph learning for open-set domain adaptation

    Yadan Luo, Zijian Wang, Zhuoxiao Chen, Zi Huang, and Mahsa Baktashmotlagh. Source-free progressive graph learning for open-set domain adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(9):11240– 11255, 2023. 3, 6

    2023

  26. [33]

    Active uni- versal domain adaptation

    Xinhong Ma, Junyu Gao, and Changsheng Xu. Active uni- versal domain adaptation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 8968– 8977, 2021. 3

    2021

  27. [34]

    Inferring latent domains for unsupervised deep domain adaptation.IEEE Transactions on Pattern Analysis and Machine Intelligence , 43(2):485–498,

    Massimiliano Mancini, Lorenzo Porzi, Samuel Rota Bulo, Barbara Caputo, and Elisa Ricci. Inferring latent domains for unsupervised deep domain adaptation.IEEE Transactions on Pattern Analysis and Machine Intelligence , 43(2):485–498,

  28. [35]

    A multi-message passing framework based on heterogeneous graphs in conversational emotion recognition

    Tao Meng, Yuntao Shou, Wei Ai, Jiayi Du, Haiyan Liu, and Keqin Li. A multi-message passing framework based on heterogeneous graphs in conversational emotion recognition. Neurocomputing, 569:127109, 2024. 2

    2024

  29. [36]

    Deep imbalanced learning for multimodal emotion recogni- tion in conversations

    Tao Meng, Yuntao Shou, Wei Ai, Nan Yin, and Keqin Li. Deep imbalanced learning for multimodal emotion recogni- tion in conversations. IEEE Transactions on Artificial Intel- ligence, 2024. 2

    2024

  30. [37]

    Revisiting multimodal emotion recognition in conversation from the perspective of graph spectrum

    Tao Meng, Fuchen Zhang, Yuntao Shou, Wei Ai, Nan Yin, and Keqin Li. Revisiting multimodal emotion recognition in conversation from the perspective of graph spectrum. arXiv preprint arXiv:2404.17862, 2024. 3

    2024 arXiv

  31. [38]

    Masked graph learning with recurrent alignment for multimodal emotion recognition in conversa- tion

    Tao Meng, Fuchen Zhang, Yuntao Shou, Hongen Shao, Wei Ai, and Keqin Li. Masked graph learning with recurrent alignment for multimodal emotion recognition in conversa- tion. IEEE/ACM Transactions on Audio, Speech, and Lan- guage Processing, 2024. 3

    2024

  32. [39]

    Co-pilot: Dynamic top-down point cloud with conditional neighborhood aggregation for multi-gigapixel histopathology image representation

    Ramin Nakhli, Allen Zhang, Ali Mirabadi, Katherine Rich, Maryam Asadi, Blake Gilks, Hossein Farahani, and Ali Bashashati. Co-pilot: Dynamic top-down point cloud with conditional neighborhood aggregation for multi-gigapixel histopathology image representation. In Proceedings of...

    2023

  33. [40]

    Dynamic changes and prognostic value of pulmonary congestion by lung ultra- sound in acute and chronic heart failure: a systematic review

    Elke Platz, Allison A Merz, Pardeep S Jhund, Ali Vazir, Ross Campbell, and John J McMurray. Dynamic changes and prognostic value of pulmonary congestion by lung ultra- sound in acute and chronic heart failure: a systematic review. European journal of heart failure, 19(9):1154–...

    2017

  34. [41]

    Assessing treatment response to oral drugs for multiple sclerosis in real-world setting: a magnims study

    Serena Ruggieri, Luca Prosperini, Sarmad Al-Araji, Pietro Osvaldo Annovazzi, Alvino Bisecco, Olga Ciccarelli, Nicola De Stefano, Massimo Filippi, Vinzenz Fleischer, Nikos Evangelou, et al. Assessing treatment response to oral drugs for multiple sclerosis in real-world setting:...

    2024

  35. [42]

    Maximum classifier discrepancy for unsuper- vised domain adaptation

    Kuniaki Saito, Kohei Watanabe, Yoshitaka Ushiku, and Tat- suya Harada. Maximum classifier discrepancy for unsuper- vised domain adaptation. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern Recognition, pages 3723–3732, 2018. 3

    2018

  36. [43]

    Transmil: Transformer based 10 correlated multiple instance learning for whole slide image classification

    Zhuchen Shao, Hao Bian, Yang Chen, Yifeng Wang, Jian Zhang, Xiangyang Ji, et al. Transmil: Transformer based 10 correlated multiple instance learning for whole slide image classification. Advances in neural information processing systems, 34:2136–2147, 2021. 7

    2021

  37. [44]

    Wasser- stein distance guided representation learning for domain adaptation

    Jian Shen, Yanru Qu, Weinan Zhang, and Yong Yu. Wasser- stein distance guided representation learning for domain adaptation. In Proceedings of the AAAI conference on ar- tificial intelligence, 2018. 3

    2018

  38. [45]

    Object detection in medical images based on hierarchical transformer and mask mechanism

    Yuntao Shou, Tao Meng, Wei Ai, Canhao Xie, Haiyan Liu, and Yina Wang. Object detection in medical images based on hierarchical transformer and mask mechanism. Com- putational Intelligence and Neuroscience, 2022(1):5863782,

    2022

  39. [46]

    Conversational emotion recognition studies based on graph convolutional neural networks and a dependent syn- tactic analysis

    Yuntao Shou, Tao Meng, Wei Ai, Sihan Yang, and Keqin Li. Conversational emotion recognition studies based on graph convolutional neural networks and a dependent syn- tactic analysis. Neurocomputing, 501:629–639, 2022. 2

    2022

  40. [47]

    Czl-ciae: Clip-driven zero-shot learning for correcting inverse age es- timation

    Yuntao Shou, Wei Ai, Tao Meng, and Keqin Li. Czl-ciae: Clip-driven zero-shot learning for correcting inverse age es- timation. arXiv preprint arXiv:2312.01758, 2023. 3

    2023 arXiv

  41. [48]

    Graph infor- mation bottleneck for remote sensing segmentation

    Yuntao Shou, Wei Ai, Tao Meng, and Nan Yin. Graph infor- mation bottleneck for remote sensing segmentation. arXiv preprint arXiv:2312.02545, 2023. 3

    2023

  42. [49]

    Graphunet: Graph make strong encoders for remote sens- ing segmentation

    YunTao Shou, Wei Ai, Tao Meng, FuChen Zhang, and KeQin Li. Graphunet: Graph make strong encoders for remote sens- ing segmentation. In 2023 IEEE 29th International Confer- ence on Parallel and Distributed Systems (ICPADS) , pages 2734–2737. IEEE, 2023. 3

    2023

  43. [50]

    Adversarial representation with intra-modal and inter-modal graph contrastive learning for multimodal emotion recogni- tion

    Yuntao Shou, Tao Meng, Wei Ai, Nan Yin, and Keqin Li. Adversarial representation with intra-modal and inter-modal graph contrastive learning for multimodal emotion recogni- tion. arXiv preprint arXiv:2312.16778, 2023. 2

    2023 arXiv

  44. [51]

    A comprehensive survey on multi-modal conversational emotion recognition with deep learning

    Yuntao Shou, Tao Meng, Wei Ai, Nan Yin, and Keqin Li. A comprehensive survey on multi-modal conversational emotion recognition with deep learning. arXiv preprint arXiv:2312.05735, 2023. 2

    2023

  45. [52]

    Efficient long-distance latent relation-aware graph neu- ral network for multi-modal emotion recognition in conver- sations

    Yuntao Shou, Wei Ai, Jiayi Du, Tao Meng, and Haiyan Liu. Efficient long-distance latent relation-aware graph neu- ral network for multi-modal emotion recognition in conver- sations. arXiv preprint arXiv:2407.00119, 2024. 2

    2024 arXiv

  46. [53]

    Spegcl: Self-supervised graph spectrum contrastive learning without positive samples

    Yuntao Shou, Xiangyong Cao, and Deyu Meng. Spegcl: Self-supervised graph spectrum contrastive learning without positive samples. arXiv preprint arXiv:2410.10365, 2024

    2024 arXiv

  47. [54]

    Contrastive graph representation learning with adversarial cross-view reconstruction and information bottleneck

    Yuntao Shou, Haozhi Lan, and Xiangyong Cao. Contrastive graph representation learning with adversarial cross-view reconstruction and information bottleneck. arXiv preprint arXiv:2408.00295, 2024. 2

    2024 arXiv

  48. [55]

    A low-rank matching attention based cross-modal feature fusion method for conversational emotion recogni- tion

    Yuntao Shou, Huan Liu, Xiangyong Cao, Deyu Meng, and Bo Dong. A low-rank matching attention based cross-modal feature fusion method for conversational emotion recogni- tion. IEEE Transactions on Affective Computing, 2024. 3

    2024

  49. [56]

    Adversarial alignment and graph fusion via information bottleneck for multimodal emotion recognition in conversations

    Yuntao Shou, Tao Meng, Wei Ai, Fuchen Zhang, Nan Yin, and Keqin Li. Adversarial alignment and graph fusion via information bottleneck for multimodal emotion recognition in conversations. Information Fusion, 112:102590, 2024. 2

    2024

  50. [57]

    Revisiting multi-modal emotion learning with broad state space models and probability-guidance fusion

    Yuntao Shou, Tao Meng, Fuchen Zhang, Nan Yin, and Ke- qin Li. Revisiting multi-modal emotion learning with broad state space models and probability-guidance fusion. arXiv preprint arXiv:2404.17858, 2024. 3

    2024 arXiv

  51. [58]

    Masked contrastive graph representation learning for age es- timation

    Yuntao Shou, Xiangyong Cao, Huan Liu, and Deyu Meng. Masked contrastive graph representation learning for age es- timation. Pattern Recognition, 158:110974, 2025. 2

    2025

  52. [59]

    Clda: Contrastive learning for semi-supervised domain adaptation

    Ankit Singh. Clda: Contrastive learning for semi-supervised domain adaptation. Advances in Neural Information Pro- cessing Systems, 34:5089–5101, 2021. 3

    2021

  53. [60]

    Can we measure long- term treatment effects in multiple sclerosis? Nature Reviews Neurology, 11(3):176–182, 2015

    Maria Pia Sormani and Paolo Bruzzi. Can we measure long- term treatment effects in multiple sclerosis? Nature Reviews Neurology, 11(3):176–182, 2015. 1

    2015

  54. [61]

    Deep neural network models for computational histopathology: A survey

    Chetan L Srinidhi, Ozan Ciga, and Anne L Martel. Deep neural network models for computational histopathology: A survey. Medical image analysis, 67:101813, 2021. 2

    2021

  55. [62]

    Gppt: Graph pre-training and prompt tuning to generalize graph neural networks

    Mingchen Sun, Kaixiong Zhou, Xin He, Ying Wang, and Xin Wang. Gppt: Graph pre-training and prompt tuning to generalize graph neural networks. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 1717–1727, 2022. 4

    2022

  56. [63]

    Feature re-embedding: Towards foun- dation model-level performance in computational pathology

    Wenhao Tang, Fengtao Zhou, Sheng Huang, Xiang Zhu, Yi Zhang, and Bo Liu. Feature re-embedding: Towards foun- dation model-level performance in computational pathology. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11343–11352, 2024. 7

    2024

  57. [64]

    Long-term cancer survival prediction using multimodal deep learning

    Lu ´ıs A Vale-Silva and Karl Rohr. Long-term cancer survival prediction using multimodal deep learning. Scientific Re- ports, 11(1):13505, 2021. 1, 2

    2021

  58. [65]

    Dual-stream multi-dependency graph neural net- work enables precise cancer survival analysis

    Zhikang Wang, Jiani Ma, Qian Gao, Chris Bain, Seiya Imoto, Pietro Li `o, Hongmin Cai, Hao Chen, and Jiangn- ing Song. Dual-stream multi-dependency graph neural net- work enables precise cancer survival analysis. Medical Im- age Analysis, 97:103252, 2024. 2, 3

    2024

  59. [66]

    The cancer genome atlas pan-cancer analysis project

    John N Weinstein, Eric A Collisson, Gordon B Mills, Kenna R Shaw, Brad A Ozenberger, Kyle Ellrott, Ilya Shmulevich, Chris Sander, and Joshua M Stuart. The cancer genome atlas pan-cancer analysis project. Nature genetics, 45(10):1113–1120, 2013. 6

    2013

  60. [67]

    Attraction and repulsion: Unsupervised domain adaptive graph contrastive learning network

    Man Wu, Shirui Pan, and Xingquan Zhu. Attraction and repulsion: Unsupervised domain adaptive graph contrastive learning network. IEEE Transactions on Emerging Topics in Computational Intelligence, 6(5):1079–1091, 2022. 3

    2022

  61. [68]

    Handling distribution shifts on graphs: An invariance per- spective

    Qitian Wu, Hengrui Zhang, Junchi Yan, and David Wipf. Handling distribution shifts on graphs: An invariance per- spective. In International Conference on Learning Repre- sentations, 2022. 3

    2022

  62. [69]

    Mome: Mixture of multimodal experts for cancer survival prediction

    Conghao Xiong, Hao Chen, Hao Zheng, Dong Wei, Yefeng Zheng, Joseph JY Sung, and Irwin King. Mome: Mixture of multimodal experts for cancer survival prediction. In In- ternational Conference on Medical Image Computing and Computer-Assisted Intervention , pages 318–328. Springer,

  63. [70]

    Heterogeneous graph attention network for unsupervised multiple-target domain adaptation

    Xu Yang, Cheng Deng, Tongliang Liu, and Dacheng Tao. Heterogeneous graph attention network for unsupervised multiple-target domain adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence , 44(4):1992– 2003, 2020. 3

    1992

  64. [71]

    From local structures to size generalization 11 in graph neural networks

    Gilad Yehudai, Ethan Fetaya, Eli Meirom, Gal Chechik, and Haggai Maron. From local structures to size generalization 11 in graph neural networks. In International Conference on Machine Learning, pages 11975–11986. PMLR, 2021. 3, 4, 5

    2021

  65. [72]

    Coco: A cou- pled contrastive framework for unsupervised domain adap- tive graph classification

    Nan Yin, Li Shen, Mengzhu Wang, Long Lan, Zeyu Ma, Chong Chen, Xian-Sheng Hua, and Xiao Luo. Coco: A cou- pled contrastive framework for unsupervised domain adap- tive graph classification. In International Conference on Ma- chine Learning, pages 40040–40053. PMLR, 2023. 3, 5, 6

    2023

  66. [74]

    Prediction model of dow jones index based on lstm-adaboost

    RunKai Ying, Yuntao Shou, and Chang Liu. Prediction model of dow jones index based on lstm-adaboost. In 2021 International Conference on Communications, Information System and Computer Engineering (CISCE), pages 808–812. IEEE, 2021. 2

    2021

  67. [75]

    Motor evoked po- tentials for multiple sclerosis, a multiyear follow-up dataset

    Jan Yperman, Veronica Popescu, Bart Van Wijmeersch, Thijs Becker, and Liesbet M Peeters. Motor evoked po- tentials for multiple sclerosis, a multiyear follow-up dataset. Scientific Data, 9(1):207, 2022. 2

    2022

  68. [76]

    A dynamic prediction model for prognosis of acute-on-chronic liver failure based on the trend of clinical indicators

    Zhenjun Yu, Yu Zhang, Yingying Cao, Manman Xu, Shaoli You, Yu Chen, Bing Zhu, Ming Kong, Fangjiao Song, Shao- jie Xin, et al. A dynamic prediction model for prognosis of acute-on-chronic liver failure based on the trend of clinical indicators. Scientific reports, 11(1):1810, 2...

    2021

  69. [77]

    A multi-view mask contrastive learning graph convolu- tional neural network for age estimation

    Yiping Zhang, Yuntao Shou, Tao Meng, Wei Ai, and Keqin Li. A multi-view mask contrastive learning graph convolu- tional neural network for age estimation. Knowledge and Information Systems, pages 1–26, 2024. 3

    2024

  70. [78]

    Predicting lymph node metastasis using histopathological images based on multiple instance learn- ing with deep graph convolution

    Yu Zhao, Fan Yang, Yuqi Fang, Hailing Liu, Niyun Zhou, Jun Zhang, Jiarui Sun, Sen Yang, Bjoern Menze, Xin- juan Fan, et al. Predicting lymph node metastasis using histopathological images based on multiple instance learn- ing with deep graph convolution. In Proceedings of the ...

    2020

  71. [79]

    Survival analysis for multimode ablation us- ing self-adapted deep learning network based on multisource features

    Ziqi Zhao, Wentao Li, Ping Liu, Aili Zhang, Jianqi Sun, and Lisa X Xu. Survival analysis for multimode ablation us- ing self-adapted deep learning network based on multisource features. IEEE Journal of Biomedical and Health Informat- ics, 28(1):19–30, 2023. 2, 3

    2023

  72. [80]

    Cross-modal translation and alignment for survival analysis

    Fengtao Zhou and Hao Chen. Cross-modal translation and alignment for survival analysis. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 21485–21494, 2023. 3, 7

    2023

  73. [81]

    Cmib: A cross modal interac- tion with information bottleneck framework for multimodal survival analysis

    Lai Zhouyou, Long Lifan, Cui Jiaqi, Xiong Deng, Liu Yuan- jun, Yin Hao, and Wang Yan. Cmib: A cross modal interac- tion with information bottleneck framework for multimodal survival analysis. In 2024 IEEE International Conference on Cybernetics and Intelligent Systems (CIS) an...

    2024

  74. [82]

    Transfer learning of graph neural networks with ego-graph information maximization.Advances in Neu- ral Information Processing Systems, 34:1766–1779, 2021

    Qi Zhu, Carl Yang, Yidan Xu, Haonan Wang, Chao Zhang, and Jiawei Han. Transfer learning of graph neural networks with ego-graph information maximization.Advances in Neu- ral Information Processing Systems, 34:1766–1779, 2021. 3 12

    2021

Pith tools

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