- Rating
- 3(+1)
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper addresses the challenges of inefficient memory updates and non-stationary dynamics in EEG-based graph learning. To address these issues, the authors propose RAGMO-Net, a regime-aware graph memory optimization framework. Specifically, it introduces a scalar-gated affine memory mechanism for efficient temporal modeling and a regime-aware proximal refinement strategy for stable memory adaptation. The authors further provide theoretical analyses of memory stability and convergence properties. Experimental results on EEG and non-EEG datasets demonstrate improved classification performance and computational efficiency compared with existing graph learning methods.
Critical strengths
-
Meaningful research problem. The paper addresses the challenges of non-stationary graph dynamics and computationally expensive memory updates, which are relevant to EEG analysis and dynamic graph learning.
-
Efficient memory architecture. The proposed scalar-gated affine memory mechanism simplifies temporal state transitions and reduces computational overhead compared with conventional recurrent architectures.
-
Extensive experimental evaluation. Experiments across EEG and non-EEG datasets, together with ablation studies and runtime comparisons, demonstrate the effectiveness and computational efficiency of the proposed framework.
Critical weaknesses
-
The scalar-gated affine update resembles conventional gated interpolation mechanisms. The authors should further clarify its methodological novelty compared with existing gated recurrent models and adaptive memory architectures.
-
Although the model introduces a learned regime score, it remains unclear whether the detected regimes correspond to meaningful changes in EEG connectivity. Further quantitative validation of regime detection would strengthen the proposed framework.
-
The stability and contraction results rely on bounded representations and additional Lipschitz-related assumptions. The authors should clarify whether these conditions hold in practice and discuss their implications for model behavior.
Questions for authors
-
The analysis mainly focuses on regime scores and connectivity visualizations. Further investigation of their relationships with established EEG biomarkers would strengthen the neurological interpretation of the learned dynamics.
-
Although subject-wise cross-validation is employed, additional cross-session or cross-dataset EEG evaluations would provide stronger evidence of robustness to non-stationary neural dynamics.
-
Figure 2 is overly simplistic and does not adequately illustrate the proposed methodology.
-
Most of the dynamic graph learning methods used for comparison were proposed more than three years ago. The authors should include more recent state-of-the-art methods as baselines to strengthen the experimental evaluation.
3 comments
Other comments (C1, C4) of reviewer ehBt
C1:Methodological Novelty of the Scalar-Gated Affine Memory Update: We agree that Eqs.(5)--(6) resemble conventional gated memory updates. The novelty is not the gate itself, but its integration into a regime-aware, optimization-consistent dynamic-graph memory. RAGMO-Net constrains and ), using a scalar retention factor rather than GRU/LSTM-style matrix-valued recurrence, reducing complexity as reflected in Table 4. Importantly, is coupled to regime score , enabling explicit adaptation to stable/changing regimes. Unlike adaptive memory architectures that rely on learned prediction--correction or long--short memory modules, RAGMO-Net applies the closed-form regime-dependent proximal refinement in Eqs.(7)--(9), combining online-gradient, Bregman-proximity, and stability terms. Lemma 1 establishes boundedness margin, while Theorem 1 gives BIBO stability and conditional contraction. Table 3 shows 19.07, 9.20, and 12.29% accuracy drops without regime score, refinement, and stability penalty, showing gains arise from their combination rather than the scalar gate alone.
C4: Regime score & Biomarkers: Tab 1 (Biomarker 1) links to frontotemporal connectivity biomarker (): median rises 0.072→0.127→0.129 across low/medium/high . High- states are more frequent in FTD vs HC (24.3% vs 8.0%) and distinguish FTD/HC (Mann--Whitney ; AUROC=0.863, ). High-vs-Low also differs (paired Wilcoxon ). These findings are consistent with EEG work showing frontotemporal connectivity provides discriminative information for FTD [1]. Tab.2 (Biomarker 2) links higher to stronger network organization:node strength (0.499→0.634; ), clustering coff(0.332→0.492; ), and global efficiency (0.346→0.612; ), with significant Holm-corrected Low-vs-High differences. These trends indicate stronger interaction, local organization, and global integration, supporting disorder-relevant network dynamics beyond one connectivity biomarker.
Tab 1: Relationship between learned regime score () and frontotemporal connectivity biomarker ()
| Metric | Low (<0.45) | Medium (0.45≤<0.55) | High (≥0.55) | Overall |
|---|---|---|---|---|
| A. Distribution across bins | ||||
| All snapshots, n (%) | 161 (0.1%) | 117,889 (85.5%) | 19,797 (14.4%) | 137,847 snapshots |
| FTD snapshots, n (%) | 93 (0.2%) | 40,767 (75.6%) | 13,085 (24.3%) | 53,945 FTD snapshots |
| HC snapshots, n (%) | 68 (0.1%) | 77,122 (91.9%) | 6,712 (8.0%) | 83,902 HC snapshots |
| B. Frontotemporal connectivity biomarker | ||||
| , median [IQR] | 0.432 [0.431, 0.437] | 0.511 [0.508, 0.515] | 0.611 [0.600, 0.615] | — |
| ), median [IQR] | 0.072 [0.051, 0.088] | 0.127 [0.117, 0.138] | 0.129 [0.115, 0.137] | — |
| FTD ), median [IQR] | 0.087 [0.070, 0.092] | 0.136 [0.124, 0.139] | 0.134 [0.122, 0.138] | High- FTD vs. HC: |
| HC ), median [IQR] | 0.051 [0.036, 0.065] | 0.122 [0.115, 0.132] | 0.107 [0.075, 0.107] | High- AUROC = 0.863 |
| C. Relationship with learned regime dynamics | ||||
| High- FTD vs. HC | Subject-level frontotemporal connectivity | — | — | AUROC = 0.863; |
| High- vs. Low- | Paired subject-level frontotemporal connectivity | — | — |
Tab 2: Relationship between learned regime score and graph-network biomarkers
| Metric | Low (<0.45) | Medium (0.45≤<0.55) | High (≥0.55) | Overall |
|---|---|---|---|---|
| A. Distribution across bins | ||||
| All valid snapshots, n (%) | 161 (0.1%) | 117,889 (85.5%) | 19,797 (14.4%) | 137,847 snapshots |
| FTD snapshots, n (%) | 93 (0.2%) | 40,767 (75.6%) | 13,085 (24.3%) | 53,945 FTD snapshots |
| HC snapshots, n (%) | 68 (0.1%) | 77,122 (91.9%) | 6,712 (8.0%) | 83,902 HC snapshots |
| B. Graph-network biomarkers | ||||
| , median [IQR] | 0.432 [0.431, 0.437] | 0.511 [0.508, 0.515] | 0.611 [0.600, 0.615] | — |
| Mean node strength | 0.499 [0.479, 0.515] | 0.612 [0.573, 0.659] | 0.634 [0.589, 0.677] | Friedman |
| Weighted clustering coefficient | 0.332 [0.279, 0.387] | 0.429 [0.384, 0.452] | 0.492 [0.429, 0.521] | Friedman |
| Global efficiency | 0.346 [0.284, 0.391] | 0.501 [0.482, 0.545] | 0.612 [0.581, 0.642] | Friedman |
| C. Low- vs. High- | ||||
| Mean node strength | Paired subject-level comparison | — | — | $$p_{Holm} |
| Weighted clustering coefficient | DO | — | — | $$p_{Holm} |
| Global efficiency | DO | — | — | $$p_{Holm} |
[1]: Yuan Ma et.al. Classification of alzheimer’s disease and frontotemporal dementia using electroencephalography to quantify communication between electrode pairs. Diagnostics, 2024
Thank you for your reply. This has addressed my concerns. I will improve my score.
Comments (C2,C3,C5,C6,C7) of Reviewer ehBt
Thanks for the comment.
C2:Quantitative validation of regime detection & EEG connectivity: Fig. 3 (main paper) links high qt to frontal/temporal connectivity alterations (sec. 5.1). Tab 1 shows qt rises from 0.572 pre-switch to 0.699 at transition and returns to 0.574 post-switch; $$\Deltaq=0.118[0.067,0.173], one-sided paired Wilcoxon p<0.001, with AUROC 0.892. Memory movement increases with qt (0.392→0.441→0.503).
Tab 1:Quantitative validation of learned regime score (qt) on DS4 dataset. Pre-switch:resting state, transition:first mental-arithmetic (MA) snapshot, and post-switch: the remaining MA snapshots. Values are summarized within subject and reported as median [IQR] across 36 paired subjects. (Δ q=qtransition−21(qpre+qpost))
C3 :Stability and Contraction: A1 (in proof) holds for fixed parameters but not uniformly during training; ut is unnormalized. A2--A3 hold for Lipschitz MLPs. Boundedness prevents memory blow-up without κ<1, while contraction requires uncertified 2KLα<δ. Fig.4 shows stabilization of refinement terms and objective across GFP peaks, not direct contraction.
C5 Cross dataset/run EEG experiment: As DS3 has multiple runs within one session, we used strict leave-one-run-out evaluation. Tab 2 shows robust DS3 cross-run and DS7→DS8/DS9 zero-shot performance under temporal and cross-dataset shifts.
Tab 2: Cross-run/dataset robustness evaluation.
C6: Fig.2 issue We need to clarify that Fig. 1 illustrates the proposed framework, and Fig. 2 presents the regime-aware memory analysis.
C7: Comparison with SOTA methods: Tab 3 compares RAGMO-Net with SALoM, EvoBrain, and FreeDyG. RAGMO-Net achieves the best Accuracy and F1 on EEG(Exp DS, DS3) and non-EEG datasets (DS7,DS8), confirming competitiveness against recent dynamic graph methods.
Tab 3: Comparison with SOTA methods
[1]:Hanwen Liu, et.al. Salom: Structure aware temporal graph networks with long-short memory updater. Advances in Neural Information Processing Systems, 2026.
[2]: Rikuto Kotoge, et.al. Evobrain: Dynamic multi-channel eeg graph modeling for time-evolving brain networks. Advances in Neural Information Processing Systems,2026
[3]:Yuxing Tian, et.al. Freedyg: Frequency enhanced continuous-time dynamic graph model for link prediction. In International Conference on Learning Representations, 2024