Pith. sign in

REVIEW 4 major objections 4 minor 66 references

Jointly modeling three functional-connectivity views with two labeled source sites and adapting to unlabeled target sites improves cross-site MDD identification, reaching 73.60% mean accuracy and 71.90% AUC over seven sites.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-03 05:08 UTC pith:L5HNDT35

load-bearing objection A coherent new architecture for multi-view multi-source MDD adaptation whose central accuracy claim needs error bars and full hyperparameter disclosure before it can be believed. the 4 major comments →

arxiv 2607.29531 v1 pith:L5HNDT35 submitted 2026-07-31 cs.CV q-bio.NC

Multi-Source Multi-View Graph Domain Adaptation with Hyperbolic Residual Encoding for Cross-Site MDD Identification from rs-fMRI

Zhanpeng Zheng , Xiran Chen , Haiteng Jiang , Renjie Tian , Qinyu Cai , Jiexi Liu , Xiaofeng Chen , Weikai Li
show 1 more author
Yansu Wang
This is my paper
classification cs.CV q-bio.NC
keywords domain adaptationmulti-view learningfunctional connectivityrs-fMRImajor depressive disordergraph attention networkshyperbolic embeddingCauchy–Schwarz divergence
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper tries to establish that cross-site identification of major depressive disorder from resting-state fMRI can be improved by treating the problem as a multi-source, multi-view, unsupervised domain-adaptation task instead of handling multi-view connectome learning and cross-site adaptation separately. It builds three complementary functional-connectivity graphs—Pearson correlation, sparse representation, and Granger causality—encodes each with its own graph attention network, fuses them through pairwise cross-view interactions and a sample-adaptive dual stream, and refines the fused representation in a Poincaré-ball hyperbolic residual encoder. Class-wise Cauchy–Schwarz alignment then reduces discrepancies between the two labeled source sites and between each source and confident pseudo-labeled target samples, alongside adversarial, information-maximization, and pseudo-label objectives. Across seven independently adapted unlabeled target sites, the framework reports a macro-averaged 73.60% accuracy and 71.90% AUC, surpassing the strongest evaluated baseline by 5.93 and 4.75 percentage points. A sympathetic reader would take this as evidence that heterogeneous connectivity views and source identities should be preserved and aligned jointly rather than pooled or adapted independently.

Core claim

On its own terms, the paper's central claim is that a unified architecture can jointly handle inter-site distribution shifts, heterogeneous functional-connectivity views, and source-to-source differences. The authors construct three view-specific graph attention encoders, form pairwise cross-view interaction features, and combine them with the original view embeddings through a sample-adaptive dual-stream fusion. A post-fusion hyperbolic residual encoder, constrained to the Poincaré ball and read out through tangent-space maps, refines the fused subject-level representation. The adaptation objective couples class-wise Cauchy–Schwarz divergence between the two sources and between each source

What carries the argument

The load-bearing machinery is class-wise Cauchy–Schwarz alignment. The Cauchy–Schwarz divergence, computed with an RBF kernel, measures distribution discrepancy between feature sets; applied per class across the two source sites (Eq. 16) and between each source and confident target pseudo-label groups (Eq. 18), it is the mechanism that preserves class structure while pulling domains together. This is carried by the two-stage optimization, where labeled sources are pretrained and then the unlabeled target participates in adaptation; the confidence-aware pseudo-label loss provides the class-wise target groupings. The post-fusion hyperbolic residual encoder (Poincaré-ball constrained, with expo

Load-bearing premise

The framework assumes that confident target predictions used as pseudo-labels are sufficiently accurate to guide class-wise alignment; if early pseudo-labels are biased by source-site distribution, the alignment and pseudo-supervision loop can reinforce that bias rather than correct it.

What would settle it

Hold out one entire acquisition site: run the same two-stage adaptation with the other seven sites as sources/targets and test on the held-out site with no adaptation on it. If accuracy on the held-out site falls to the level of the no-adaptation baseline (around 65.6% ACC), the reported gains come from transductive use of target data, not from transferable alignment.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • If the framework generalizes as reported, a new acquisition site's unlabeled scans could be used to adapt an MDD classifier without manual labeling at that site.
  • Preserving the two source sites as distinct sources outperformed pooling them (73.60% vs. 70.12% for the stronger single source), implying that source identity should be kept rather than collapsed.
  • Removing the domain-adaptation objective drops mean ACC from 73.60% to 65.61% and AUC from 71.90% to 52.58%, indicating the alignment terms, not the architecture alone, drive transfer.
  • Removing the hyperbolic encoder drops ACC to 64.86%, and the best single-view variant reaches only 70.34%, so multi-view fusion and curvature-aware refinement each contribute.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • My inference: the reported evaluation is transductive—each target site's unlabeled data participates in adaptation—so the numbers are not a direct measure of how the model would behave on a brand-new site never seen during adaptation; an independent-site holdout test (train/adapt on seven sites, test on an eighth) would be a stricter check.
  • My inference: the confidence-aware pseudo-label loop could suffer from confirmation bias if early pseudo-labels inherit source-site biases; a useful stress test would be to corrupt a fraction of confident pseudo-labels or vary the threshold schedule and observe whether accuracy degrades smoothly or abruptly.
  • My inference: because target sites vary widely in size and class balance, class-wise Cauchy–Schwarz estimates for small classes may be noisy; an extension would be to weight class terms by sample size or use robust kernel estimators.
  • My inference: the hyperbolic residual encoder's benefit could be disentangled from extra nonlinear capacity by swapping it for a Euclidean residual gate of matched parameter count; the paper does not report this comparison.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes a multi-source, multi-view unsupervised graph domain adaptation framework for cross-site MDD identification from rs-fMRI. Three functional-connectivity views (Pearson correlation, sparse representation, Granger causality) are encoded by separate graph attention networks; pairwise cross-view interactions are integrated through a dual-stream adaptive fusion, followed by a Poincaré-ball hyperbolic residual encoder. Adaptation is driven by class-wise Cauchy–Schwarz alignment, adversarial learning, information maximization, and confidence-aware pseudo-labeling in a two-stage protocol. On seven target sites from REST-meta-MDD, the paper reports 73.60% mean accuracy and 71.90% mean AUC, with the highest per-site ACC on all seven sites and a 5.93-point mean ACC improvement over the H2MSDA baseline. The authors also include source-setting sensitivity, single/dual-source analysis, and ablations, and they release source code.

Significance. If the reported results are stable, the paper would make a useful contribution by jointly addressing multi-source and multi-view shifts in a transductive cross-site brain-imaging setting. The architectural components — especially the explicit modeling of pairwise cross-view interactions and the post-fusion hyperbolic residual — are well motivated, and the code release is a concrete strength for reproducibility. However, the central empirical claim is currently supported by single-run numbers without variance estimates or significance tests, and several target sites have very small sample sizes. The evidence is also explicitly transductive, so the abstract's wording about 'effective generalization under heterogeneous acquisition conditions' overstates what is demonstrated. The contribution is potentially solid but needs additional experimental rigor before publication.

major comments (4)
  1. [Table 2 and Section 5.2] The headline claim rests entirely on Table 2, but the paper reports only one run per method/site, with no standard deviations, confidence intervals, or significance tests. Target sites are small (Site6 n=30, Site12 n=38), so per-site differences are fragile: the 6.67-point ACC advantage at Site6 corresponds to roughly 2 of 30 samples, and the claimed 'highest ACC on all seven sites' may be within chance variation. Please provide results over multiple random seeds (at least 5–10), report mean±std per site and macro-averaged, and include paired tests versus H2MSDA and AUF-A. This is load-bearing because the superiority claim depends on the per-site gaps being non-noise.
  2. [Equations (2), (23) and Section 5.2] Several free hyperparameters are not reported: the loss weights λ_ss, λ_st, λ_adv, λ_im, λ_pl and their schedules, the pseudo-label confidence threshold τ and its schedule, the RBF kernel bandwidth in Eq. (17), the Poincaré-ball curvature c, and the graph sparsity criterion. The text fixes λ_SR=0.01 and GCM lag order=1 but not the others. Please report all hyperparameter values and, critically, the selection protocol. If any hyperparameter was tuned using target-site labels or by inspecting target-domain performance, the transductive unsupervised claim would be violated; if it was selected on a held-out validation split, describe the split. Without this, the reported numbers are not independently reproducible.
  3. [Equations (18) and (22)] The pseudo-label loop is a potential source of confirmation bias: class-wise source–target alignment and the pseudo-label loss both use the model's own confident target predictions, and the paper provides no analysis of pseudo-label quality or a correction mechanism. Early biased predictions can be reinforced by the alignment and pseudo-supervision, inflating apparent adaptation quality. Please report pseudo-label accuracy and coverage across adaptation iterations (or equivalent diagnostics), and discuss whether the confidence threshold schedule adequately mitigates this. If it is not mitigated, an independent-site (non-transductive) evaluation would be needed to separate the loop's actual benefit from self-reinforcement.
  4. [Abstract and Section 6] The abstract states that the framework 'demonstrat[es] effective generalization under heterogeneous acquisition conditions,' but the conclusion explicitly limits the evidence to the transductive protocol: all target samples are used for adaptation, and no independent-site validation is provided. This is not an internal inconsistency, but the wording oversells the result. Please revise the abstract and Section 5.1 to say 'transductive adaptation performance' rather than generalization, or add an independent-site evaluation (e.g., train on Site20/21, adapt on a subset of one target site, and evaluate on held-out subjects from that site). The distinction matters for the clinical claim.
minor comments (4)
  1. [Equation (1)] The Pearson correlation view uses the absolute value of the correlation, but the text does not state whether sign information is intentionally discarded. Please clarify the rationale and whether the sign was tested in ablations.
  2. [Section 3.2] The Granger causality construction is described only in general terms. Specify whether bivariate or multivariate Granger causality was used, how the lag was chosen, and how the residual-variance reduction was computed. Also state whether the GCM matrix was symmetrized.
  3. [Table 5] In the ablation table, removing the hyperbolic encoder reduces AUC by 17.34 points (0.7190 vs 0.5456), which is a much larger effect than the ACC drop. A brief explanation of this asymmetry would help the reader interpret the hyperbolic module's role.
  4. [Figure 2] The t-SNE visualization is presented as qualitative evidence, which is appropriate. Please state the t-SNE parameters (perplexity, number of iterations) and, if possible, provide a reproducibility note for the exact visualization.

Circularity Check

0 steps flagged

No significant circularity: the paper reports an empirical transductive benchmark; no claimed result reduces to its own inputs or to a self-citation chain.

full rationale

The paper makes no claim to derive a first-principles result; it presents an architecture and an empirical benchmark. Target labels are used 'only after training to compute ACC and AUC' (Section 3.1), so the headline numbers are a transductive UDA evaluation, not a prediction of labels that participated in fitting. The objectives in Eqs. (16)-(23) are standard domain-adaptation and pseudo-label losses; none is defined in terms of the reported ACC/AUC. The pseudo-label loop (Eqs. 18, 21-22) uses the model's own confident target predictions to form class-wise target sets and pseudo-supervision. This is a self-training mechanism and can suffer confirmation bias, but it is not an equation-level equivalence between fitted inputs and the reported result; the paper also describes progressive confidence-threshold scheduling to limit early wrong labels. The conclusion explicitly limits evidence to 'the evaluated transductive protocol' and lists independent-site validation as future work, which is a scope limitation rather than circularity. Small-site sample sizes and the absence of error bars or significance tests are statistical-reporting concerns, not circularity. No load-bearing self-citation appears: the only author-overlapping reference is the REST-meta-MDD dataset citation (Yan et al., 2019), which is external data, not a theorem or fitted parameter. No uniqueness theorem, ansatz, or prior result from the authors is invoked to force the design. Thus no step meets the required standard of exhibiting a specific reduction to its own inputs or to a self-citation chain.

Axiom & Free-Parameter Ledger

7 free parameters · 6 axioms · 0 invented entities

The empirical claim depends on a large set of hand-set or unreported numbers (lambda_SR, lag order, sparsity, curvature, loss weights, threshold schedules, kernel bandwidth). None of these are derived; they are chosen to make the pipeline work, and their values are not shipped.

free parameters (7)
  • lambda_SR = 0.01
    Sparsity weight in Eq. (2) for the sparse-representation view; set by hand, not justified.
  • GCM lag order = 1
    Lag order for Granger causality mapping; set by hand.
  • Graph sparsity criterion = not specified
    Strongest connections retained per node by a predefined criterion; threshold not reported, directly shapes all three graphs.
  • Poincare ball curvature c = not specified
    Curvature in Eqs. (10)-(12) is described as curvature-controlled but no value or schedule is given.
  • Loss weights lambda_ss, lambda_st, lambda_adv, lambda_im, lambda_pl and schedules = not specified
    Eq. (23) weights are dynamically scheduled; values not reported, and no validation protocol is described.
  • Pseudo-label confidence threshold tau and schedule = not specified
    Eq. (21) threshold is progressively relaxed; initial value and update rule not reported.
  • RBF kernel bandwidth = not specified
    Cauchy-Schwarz divergence in Eq. (17) uses an RBF kernel; bandwidth is a free parameter but omitted.
axioms (6)
  • domain assumption AAL116 ROI time series from REST-meta-MDD are comparable across sites after z-standardization
    Section 3.2 loads released preprocessed signals; inter-site comparability is assumed rather than demonstrated.
  • domain assumption Pearson, sparse representation, and Granger causality views capture complementary disease-relevant information
    Section 2.2 motivates this; no evidence is given that these three views, rather than alternatives, are necessary or sufficient.
  • domain assumption The sparsification procedure preserves disease-relevant edges at every site
    Section 3.2 retains strongest connections per node with a predefined criterion; if the criterion discards discriminative edges, all downstream claims fail.
  • domain assumption Graph attention and hyperbolic residual encoding provide a valid inductive bias for brain hierarchy
    Sections 4.3 and 4.6 assert hierarchy; no controlled comparison with Euclidean-only encoders is isolated except the ablation.
  • standard math Cauchy-Schwarz divergence with RBF kernel is a reliable class-conditional alignment objective
    Eq. (17) uses standard divergence; validity relies on kernel bandwidth and sample sizes, with some target classes having as few as 15 samples.
  • domain assumption Source and target share the same label space and task
    Standard UDA assumption stated implicitly in Section 4.1; unverified for MDD diagnosis across sites.

pith-pipeline@v1.3.0-daily-deepseek · 10254 in / 12563 out tokens · 113937 ms · 2026-08-03T05:08:12.946848+00:00 · methodology

0 comments
read the original abstract

Cross-site identification of major depressive disorder (MDD) from resting-state functional magnetic resonance imaging (rs-fMRI) is hindered by inter-site distribution shifts and heterogeneous functional connectivity (FC) views. These views capture complementary neural relationships but exhibit distinct site biases and graph topologies, complicating alignment without sacrificing disease-relevant information or cross-view consistency. Existing studies largely treat multi-view connectome learning and cross-site adaptation separately. To the best of our knowledge, few studies have jointly modeled multiple FC views under multi-source unsupervised domain adaptation for cross-site rs-fMRI-based MDD classification. We construct Pearson correlation, sparse representation, and Granger causality graphs, each encoded by a view-specific graph attention network. Dual-stream adaptive fusion explicitly integrates pairwise cross-view interactions, followed by lightweight hyperbolic residual encoding for curvature-aware representation refinement. Class-wise Cauchy--Schwarz alignment reduces inter-source and source-target discrepancies, complemented by adversarial learning, information maximization, and confidence-aware pseudo-labeling. Across seven unlabeled target domains, our framework achieves 73.60% mean accuracy and 71.90% AUC, demonstrating effective generalization under heterogeneous acquisition conditions. These results highlight the effectiveness of unified heterogeneous-view modeling, curvature-aware refinement, and multi-source domain adaptation for cross-site MDD identification.The source code is at https://github.com/OPUS-Lightphenexx/MM-HyperGDA

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

66 extracted references

  1. [1]

    CNS Neuroscience & Therapeutics , volume=

    Identification of Major Depressive Disorder Using Multiple Functional Connection Patterns , author=. CNS Neuroscience & Therapeutics , volume=. 2026 , publisher=

    2026

  2. [2]

    Cerebral Cortex , volume=

    Towards an accurate autism spectrum disorder diagnosis: multiple connectome views from fMRI data , author=. Cerebral Cortex , volume=. 2024 , publisher=

    2024

  3. [3]

    Cerebral Cortex , volume=

    Multipattern graph convolutional network-based autism spectrum disorder identification , author=. Cerebral Cortex , volume=. 2024 , publisher=

    2024

  4. [4]

    Imaging Neuroscience , volume=

    Is Pearson’s correlation coefficient enough for functional connectivity in fMRI? , author=. Imaging Neuroscience , volume=. 2025 , publisher=

    2025

  5. [5]

    Brain connectivity , volume=

    Functional and effective connectivity: a review , author=. Brain connectivity , volume=. 2011 , publisher=

    2011

  6. [6]

    Frontiers in systems neuroscience , volume=

    A tutorial review of functional connectivity analysis methods and their interpretational pitfalls , author=. Frontiers in systems neuroscience , volume=. 2016 , publisher=

    2016

  7. [7]

    Neuroimage , volume=

    Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain , author=. Neuroimage , volume=. 2002 , publisher=

    2002

  8. [8]

    IEEE Journal of Biomedical and Health Informatics , year=

    Multi-site rs-fmri domain alignment for autism spectrum disorder auxiliary diagnosis based on hyperbolic space , author=. IEEE Journal of Biomedical and Health Informatics , year=

  9. [9]

    Neuroimage , volume=

    Statistical power and prediction accuracy in multisite resting-state fMRI connectivity , author=. Neuroimage , volume=. 2017 , publisher=

    2017

  10. [10]

    Medical image analysis , volume=

    Unsupervised cross-domain functional MRI adaptation for automated major depressive disorder identification , author=. Medical image analysis , volume=. 2023 , publisher=

    2023

  11. [11]

    Nature Reviews Disease Primers , volume=

    Major depressive disorder , author=. Nature Reviews Disease Primers , volume=. 2023 , publisher=

    2023

  12. [12]

    and Dai, W

    Kan, X. and Dai, W. and Cui, H. and Zhang, Z. and Guo, Y. and Yang, C. , title =. Advances in Neural Information Processing Systems (NeurIPS) , volume =. 2022 , url =

    2022

  13. [13]

    and Dong, G

    Luo, X. and Dong, G. and Wu, J. and Beheshti, A. and Yang, J. and Xue, S. , title =. Proceedings of the Seventeenth ACM International Conference on Web Search and Data Mining (WSDM) , pages =. 2024 , doi =

    2024

  14. [14]

    and Yang, E

    Yao, D. and Yang, E. and Guan, H. and Sui, J. and Zhang, Z. and Liu, M. , title =. Medical Image Computing and Computer Assisted Intervention (MICCAI) , series =. 2021 , doi =

    2021

  15. [15]

    and Li, C

    Luo, G. and Li, C. and Cui, H. and Sun, L. and He, L. and Yang, C. , title =. IEEE International Conference on Bioinformatics and Biomedicine (BIBM) , pages =. 2022 , doi =

    2022

  16. [16]

    Graph Attention Networks , booktitle =

    Veli. Graph Attention Networks , booktitle =. 2018 , url =

    2018

  17. [17]

    and Zhou, M

    Yang, M. and Zhou, M. and Ying, R. and Chen, Y. and King, I. , title =. Proceedings of the 40th International Conference on Machine Learning (ICML) , series =. 2023 , url =

    2023

  18. [18]

    and Wan, Z

    Mishne, G. and Wan, Z. and Wang, Y. and Yang, S. , title =. Proceedings of the 40th International Conference on Machine Learning (ICML) , series =. 2023 , url =

    2023

  19. [19]

    and Verma, H

    Yang, M. and Verma, H. and Zhang, D. C. and Liu, J. and King, I. and Ying, R. , title =. Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) , pages =. 2024 , doi =

    2024

  20. [20]

    Augmentation-Based Unsupervised Cross-Domain Functional

    Ma, Yunling and Zhang, Chaojun and Wang, Xiaochuan and Wang, Qianqian and Cao, Liang and Zhang, Limei and Liu, Mingxia , journal =. Augmentation-Based Unsupervised Cross-Domain Functional. 2026 , month = mar, doi =

    2026

  21. [21]

    Multi-Site rs-

    Luo, Yiqian and Chen, Qiurong and Li, Fali and Xu, Peng and Zhang, Yangsong , journal =. Multi-Site rs-. 2026 , month = feb, doi =

    2026

  22. [22]

    Sun, Baochen and Saenko, Kate , booktitle =. Deep. 2016 , doi =

    2016

  23. [23]

    Journal of Machine Learning Research , volume =

    Domain-Adversarial Training of Neural Networks , author =. Journal of Machine Learning Research , volume =

  24. [24]

    International Conference on Learning Representations , year =

    How Powerful Are Graph Neural Networks? , author =. International Conference on Learning Representations , year =

  25. [25]

    International Conference on Learning Representations , year =

    Semi-Supervised Classification with Graph Convolutional Networks , author =. International Conference on Learning Representations , year =

  26. [26]

    Advances in Neural Information Processing Systems , volume =

    Attention Is All You Need , author =. Advances in Neural Information Processing Systems , volume =

  27. [27]

    American Journal of Psychiatry , volume =

    Multimodal Abnormalities of Brain Structure and Function in Major Depressive Disorder: A Meta-Analysis of Neuroimaging Studies , author =. American Journal of Psychiatry , volume =. 2020 , doi =

    2020

  28. [28]

    Deep Learning for Brain Disorder Diagnosis Based on

    Yin, Wutao and Li, Longhai and Wu, Fang-Xiang , journal =. Deep Learning for Brain Disorder Diagnosis Based on. 2022 , doi =

    2022

  29. [29]

    IEEE Transactions on Medical Imaging , volume =

    A Mutual Multi-Scale Triplet Graph Convolutional Network for Classification of Brain Disorders Using Functional or Structural Connectivity , author =. IEEE Transactions on Medical Imaging , volume =. 2021 , doi =

    2021

  30. [30]

    IEEE Journal of Biomedical and Health Informatics , volume =

    Graph Autoencoders for Embedding Learning in Brain Networks and Major Depressive Disorder Identification , author =. IEEE Journal of Biomedical and Health Informatics , volume =. 2024 , month = mar, doi =

    2024

  31. [31]

    and Sivgin, Irmak and Dalmaz, Onat and Dar, Salman U

    Bedel, Hasan A. and Sivgin, Irmak and Dalmaz, Onat and Dar, Salman U. H. and. Medical Image Analysis , volume =. 2023 , doi =

    2023

  32. [32]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , volume =

    Domain Generalization: A Survey , author =. IEEE Transactions on Pattern Analysis and Machine Intelligence , volume =. 2023 , doi =

    2023

  33. [33]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages =

    Understanding and Improving Source-Free Domain Adaptation from a Theoretical Perspective , author =. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages =

  34. [34]

    Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , pages =

    Robust Source-Free Domain Adaptation for Fundus Image Segmentation , author =. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , pages =

  35. [35]

    Proceedings of Machine Learning Research , volume =

    On-the-Fly Test-Time Adaptation for Medical Image Segmentation , author =. Proceedings of Machine Learning Research , volume =

  36. [36]

    Proceedings of the 41st International Conference on Machine Learning , year =

    Test-Time Model Adaptation with Only Forward Passes , author =. Proceedings of the 41st International Conference on Machine Learning , year =

  37. [37]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages =

    Source-Free Domain Adaptation with Frozen Multimodal Foundation Model , author =. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages =

  38. [38]

    Proceedings of the National Academy of Sciences , volume =

    Reduced Default Mode Network Functional Connectivity in Patients with Recurrent Major Depressive Disorder , author =. Proceedings of the National Academy of Sciences , volume =. 2019 , doi =

    2019

  39. [39]

    2010 , doi =

    Yan, Chao-Gan and Zang, Yu-Feng , journal =. 2010 , doi =

    2010

  40. [40]

    and Barnes, Kelly A

    Power, Jonathan D. and Barnes, Kelly A. and Snyder, Abraham Z. and Schlaggar, Bradley L. and Petersen, Steven E. , journal =. Spurious but Systematic Correlations in Functional Connectivity. 2012 , doi =

    2012

  41. [41]

    and Hwang, Kai and Luna, Beatriz , journal =

    Hallquist, Michael N. and Hwang, Kai and Luna, Beatriz , journal =. The Nuisance of Nuisance Regression: Spectral Misspecification in a Common Approach to Resting-State. 2013 , doi =

    2013

  42. [42]

    Journal of the Royal Statistical Society: Series B (Methodological) , volume =

    Regression Shrinkage and Selection via the Lasso , author =. Journal of the Royal Statistical Society: Series B (Methodological) , volume =. 1996 , doi =

    1996

  43. [43]

    Econometrica , volume =

    Investigating Causal Relations by Econometric Models and Cross-Spectral Methods , author =. Econometrica , volume =. 1969 , doi =

    1969

  44. [44]

    Journal of Neuroscience , volume =

    Granger Causality Analysis in Neuroscience and Neuroimaging , author =. Journal of Neuroscience , volume =. 2015 , doi =

    2015

  45. [45]

    Advances in Neural Information Processing Systems , volume =

    Deep Sets , author =. Advances in Neural Information Processing Systems , volume =

  46. [46]

    Nickel, Maximilian and Kiela, Douwe , booktitle =. Poincar

  47. [47]

    Advances in Neural Information Processing Systems , volume =

    Hyperbolic Neural Networks , author =. Advances in Neural Information Processing Systems , volume =

  48. [48]

    Advances in Neural Information Processing Systems , volume =

    Hyperbolic Graph Convolutional Neural Networks , author =. Advances in Neural Information Processing Systems , volume =

  49. [49]

    Proceedings of the IEEE/CVF International Conference on Computer Vision , pages =

    Moment Matching for Multi-Source Domain Adaptation , author =. Proceedings of the IEEE/CVF International Conference on Computer Vision , pages =

  50. [50]

    Advances in Neural Information Processing Systems , volume =

    Adversarial Multiple Source Domain Adaptation , author =. Advances in Neural Information Processing Systems , volume =

  51. [51]

    Proceedings of the 32nd International Conference on Machine Learning , series =

    Learning Transferable Features with Deep Adaptation Networks , author =. Proceedings of the 32nd International Conference on Machine Learning , series =

  52. [52]

    Advances in Neural Information Processing Systems , volume =

    Conditional Adversarial Domain Adaptation , author =. Advances in Neural Information Processing Systems , volume =

  53. [53]

    Journal of Machine Learning Research , volume =

    A Kernel Two-Sample Test , author =. Journal of Machine Learning Research , volume =

  54. [54]

    Journal of the Franklin Institute , volume =

    The Cauchy--Schwarz Divergence and Parzen Windowing: Connections to Graph Theory and Mercer Kernels , author =. Journal of the Franklin Institute , volume =. 2006 , doi =

    2006

  55. [55]

    Advances in Neural Information Processing Systems , volume =

    Semi-Supervised Learning by Entropy Minimization , author =. Advances in Neural Information Processing Systems , volume =

  56. [56]

    ICML Workshop on Challenges in Representation Learning , year =

    Pseudo-Label: The Simple and Efficient Semi-Supervised Learning Method for Deep Neural Networks , author =. ICML Workshop on Challenges in Representation Learning , year =

  57. [57]

    Advances in Neural Information Processing Systems , volume =

    FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence , author =. Advances in Neural Information Processing Systems , volume =

  58. [58]

    Proceedings of the 37th International Conference on Machine Learning , series =

    Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation , author =. Proceedings of the 37th International Conference on Machine Learning , series =

  59. [59]

    International Conference on Learning Representations , year =

    Decoupled Weight Decay Regularization , author =. International Conference on Learning Representations , year =

  60. [60]

    Paszke, Adam and Gross, Sam and Massa, Francisco and Lerer, Adam and Bradbury, James and Chanan, Gregory and Killeen, Trevor and Lin, Zeming and Gimelshein, Natalia and Antiga, Luca and others , booktitle =

  61. [61]

    Visualizing Data Using

    van der Maaten, Laurens and Hinton, Geoffrey , journal =. Visualizing Data Using

  62. [62]

    An Introduction to

    Fawcett, Tom , journal =. An Introduction to. 2006 , doi =

    2006

  63. [63]

    Proceedings of the National Academy of Sciences , volume=

    Reduced default mode network functional connectivity in patients with recurrent major depressive disorder , author=. Proceedings of the National Academy of Sciences , volume=. 2019 , doi=

    2019

  64. [64]

    Psychoradiology , volume=

    The DIRECT consortium and the REST-meta-MDD project: Towards neuroimaging biomarkers of major depressive disorder , author=. Psychoradiology , volume=. 2022 , doi=

    2022

  65. [65]

    Cerebral Cortex , volume=

    Classification of major depressive disorder using an attention-guided unified deep convolutional neural network and individual structural covariance network , author=. Cerebral Cortex , volume=. 2023 , doi=

    2023

  66. [66]

    Human Brain Mapping , volume=

    Classification of MDD using a Transformer classifier with large-scale multisite resting-state fMRI data , author=. Human Brain Mapping , volume=. 2024 , doi=

    2024