Pith. sign in

REVIEW 1 major objections 4 minor 26 references

CAM-Guided Saliency Cutout and Image-Based Malware Classification

T0 review · 1 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read Saliency-guided cutout does not improve grayscale malware classification.

desk verdict A well-scoped negative result with unusually transparent reporting; the central claim survives the main ambiguity, but the random-cutout control's static/dynamic behavior needs checking before the paired claims can be trusted. read the letter →

arxiv 2608.11634 v1 pith:GDDPJJLG submitted 2026-08-12 cs.CV cs.LG

Yasaman Ebrahimi , Martin Jurecek , Mark Stamp This is my paper · ORCID
classification cs.CVcs.LG
keywords HiResCAMcutoutregularizationsaliency-guidedaugmentationmalwareimageclassificationRawMal-TFCIFAR-100Net18data
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper tests whether cutout regularization can be improved by placing the removed square where a saliency map says the image matters least or most. It compares no cutout, random cutout, low-saliency cutout, and high-saliency cutout on grayscale RawMal-TF malware images and CIFAR-100, all with ResNet18 and three seeds. The central result is negative for malware: the no-cutout baseline reaches $72.83\% \pm 0.16\%$ mean best validation accuracy, and every cutout condition lands lower, with high-saliency cutout generally the worst. The same pipeline on CIFAR-100 improves under low-saliency cutout, reaching $63.51\% \pm 0.36\%$ versus $62.65\% \pm 0.57\%$ for no cutout. The paper reads this as evidence that saliency-guided cutout is domain dependent and that malware images cannot be treated as natural images.

What carries the argument

The machinery is a HiResCAM (High-Resolution Class Activation Mapping) saliency heatmap produced by a no-cutout ResNet18 teacher and used to place square cutout windows. For each training image, the teacher's predicted-class heatmap is normalized and resized; candidate squares are scored by mean saliency, the lowest- or highest-scoring $10\%$ form the candidate set, and one window is chosen per copy with a copy-specific seed. The chosen windows are cached, so each sample-index and augmentation-index pair maps to the same static window every epoch. The student is then trained on the original image plus $M$ occluded copies, with validation images never cut out, and every saliency condition is compared against random cutout with identical cutout area and $M$. This design is meant to isolate the placement rule as the only difference between conditions.

What would settle it

Run the random-cutout control with exactly the same cached static window per sample and augmentation index used by the CAM-guided conditions; if the low-saliency advantage over random on CIFAR-100 shrinks or reverses, the reported placement effect is partly an artifact of mask diversity rather than of saliency.

Watch

Extended reading notes

Core claim

The paper's central claim is that HiResCAM-guided cutout does not improve grayscale malware family classification in the tested controlled setting. On RawMal-TF, the no-cutout ResNet18 baseline attains the highest mean best validation accuracy, $72.83\% \pm 0.16\%$; the best cutout condition is random cutout with $M=4$ and $30\%$ area at $71.55\% \pm 0.45\%$, and the best low-saliency condition is $71.43\% \pm 1.12\%$. Low-saliency cutout is seed-sensitive relative to matched random cutout, ranging from $-0.56$ to $+0.46$ percentage points across the eight area and $M$ cells, while high-saliency cutout is negative in most cells, with paired deficits from $-0.21$ to $-2.00$ percentage points. On CIFAR-100 the pattern reverses: low-saliency cutout with $M=4$ and $10\%$ area reaches $63.51\% \pm 0.36\%$ versus $62.65\% \pm 0.57\%$ for no cutout, and low-saliency placement protects large-mask performance, while high-saliency cutout is consistently harmful. The conclusion is that the value of saliency-guided cutout depends on the image domain, so malware-image saliency should not be equated with natural-image object saliency.

Load-bearing premise

The controlled comparison assumes that the CAM-guided and random cutout conditions differ only in where the square is placed, yet the paper specifies that CAM-guided masks are static across epochs and does not say whether the random control re-samples its window each epoch.

Editorial extensions

If this is right

  • On grayscale RawMal-TF with ResNet18 and 100 epochs, adding cutout in any tested form lowers mean best validation accuracy, so the no-cutout baseline is the strongest setting in this controlled sweep.
  • Low-saliency cutout is not reliably better than random cutout on malware: its paired advantage ranges from $-0.56$ to $+0.46$ percentage points and changes sign across seeds, while high-saliency cutout is generally worse by up to 2.00 percentage points.
  • On CIFAR-100, low-saliency cutout improves peak validation accuracy over no cutout at small mask sizes and over random cutout for large masks, whereas high-saliency cutout is worse than random in every tested cell.
  • The same implementation can produce a positive saliency-guided result, so the negative RawMal-TF outcome is a domain effect rather than a broken pipeline.

Reading between the lines

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

  • A tighter random control that caches one static window per sample and augmentation index, exactly like the CAM-guided runs, would test whether the reported low-versus-random differences are inflated by mask diversity; the paper does not state whether the random condition re-samples each epoch.
  • The teacher's HiResCAM is computed for its predicted class from the final convolutional layer, so using ground-truth labels or intermediate layers could change both low- and high-saliency rankings, an extension the paper leaves open.
  • Aligning masks with executable structure, such as PE-section boundaries or byte ranges, rather than image-saliency squares, is the paper's own suggestion for future work and could separate the failure of square saliency from the failure of cutout itself.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

1 major / 4 minor

Summary. This paper experimentally evaluates whether cutout regularization with regions selected by HiResCAM saliency improves image-based malware classification relative to no cutout and standard random cutout. Using ResNet18 with 100 epochs, seeds 42–44, cutout areas of 5%, 10%, 20%, and 30%, and M=4 and M=8 augmented copies, the authors compare four conditions (no cutout, random, low-saliency, high-saliency) on grayscale RawMal-TF images (17 families, 1,000 samples per family) and, for contrast, on CIFAR-100. The paper reports that on RawMal-TF the no-cutout baseline achieves the highest mean best validation accuracy (72.83%±0.16%) and that every cutout condition is lower; low-saliency cutout is seed-sensitive relative to matched random cutout, and high-saliency cutout is generally worse. On CIFAR-100, low-saliency cutout at 10% area improves over the no-cutout baseline (63.51% vs. 62.65%) while high-saliency cutout is consistently harmful. The authors conclude that the value of saliency-guided cutout is domain dependent and that malware images should not be treated as equivalent to natural images.

Significance. If the experimental comparison is sound, this paper provides a useful controlled negative result for CAM-guided cutout in malware classification and a contrasting positive result on CIFAR-100. The study's strengths include within-seed paired comparisons, multiple seeds and cutout areas, a teacher–student setup that avoids circularity (the teacher is trained without cutout and the student is trained on different augmented data), and explicit acknowledgment that three seeds make the statistics exploratory. The public repository and per-run metrics files support reproducibility. The main caveat is that the central matched comparison between saliency-guided and random cutout presumes that the two conditions differ only in how the cutout window is selected; if the random control re-samples masks each epoch while the CAM-guided masks are static, the measured differences conflate placement rule with mask diversity. This issue must be resolved before the secondary conclusions about low- versus high-saliency placement and the CIFAR-100 advantage over random cutout can be considered reliable.

major comments (1)
  1. [Section 3.3, Tables 3 and 4] The paper does not state whether the standard random cutout control also uses a static window per sample/copy index or re-samples a new window each epoch. Section 3.3 explicitly says that for CAM-guided cutout "the augmented copies are static: a sample index and augmentation index map to the same selected window in every epoch," but no corresponding statement is given for the random condition. If random cutout re-samples each epoch, then the random control enjoys greater mask diversity over the 100 training epochs, meaning the paired differences in Tables 3 and 4 (low-minus-random and high-minus-random) do not isolate the effect of saliency-based placement; they conflate the placement rule with temporal mask diversity. This directly affects the paper's secondary claims that low-saliency cutout is seed-sensitive relative to random on RawMal-TF, that high-saliency cutout is generally harmful, and that low-saliency cutout improves over random on CIFAR-100. The primary negative claim that no-cutout beats all cutout conditions is less affected, but the controlled comparison that motivates the study is compromised. The authors should state explicitly how the random control is implemented; if it is dynamic, they should either re-run with matched static masks or substantially temper these comparative conclusions.
minor comments (4)
  1. [Section 1, Introduction] In the concluding sentence of the introduction, "the field fo malware analysis" should read "the field of malware analysis."
  2. [Section 2.1] The sentence "a model becomes more stronger" contains a redundant comparative; it should be "a model becomes stronger."
  3. [Section 3.3] The description of candidate-window selection says that a "specified, copy-specific seed value" is used, but it does not clarify whether this seed is also epoch-specific or fixed across epochs. This is closely related to the major issue about static versus dynamic masks and should be clarified in the revision.
  4. [Section 5.5] The paper appropriately acknowledges that only validation metrics are available and that no held-out test set is reported. Given this limitation, the abstract and conclusions should perhaps more consistently use the phrase "validation accuracy" rather than "accuracy" to avoid any impression of test-set generalization, though the current phrasing is mostly careful.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper reports controlled experiments whose target outcomes are measured, not derived from their inputs.

full rationale

The paper is an empirical controlled study rather than a derivation, and its central claims are measured outcomes rather than consequences of construction. The teacher model is a no-cutout ResNet18 checkpoint trained on the original training images, and the student models are trained on cutout-augmented copies and evaluated on unmasked validation data, so the target result (whether saliency-guided cutout improves accuracy) is not an input to the method. The low-saliency and high-saliency conditions select masks from HiResCAM maps of the teacher's predicted class, while the random condition selects masks without saliency, and the paired comparisons in Tables 3 and 4 are computed within matched dataset, seed, area, and augmentation multiplicity settings. None of the reported accuracies are fitted parameters renamed as predictions, and no equation in the paper reduces the output to an input by definition. The references to RawMal-TF [2,3] and to the repository [9] supply the dataset, code, and prior transformation comparisons, but they do not carry the load of the conclusion that cutout fails to improve grayscale malware classification; that conclusion is stated as a negative experimental finding. The paper also explicitly limits its claims to validation metrics and acknowledges missing test-set and per-family metrics, which is the opposite of presenting a forced result as a derivation. The static-cache detail in Section 3.3 ('The augmented copies are static: a sample index and augmentation index map to the same selected window in every epoch') raises a possible experimental confound if the random control re-samples masks each epoch, because the random condition would then have greater mask diversity; however, that is a methodological concern about control equivalence, not a circularity in which the conclusion is equivalent to the input by construction. Accordingly, no circular step can be exhibited, and the appropriate score is 0.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The study is empirical; it does not derive constants or invent entities. The hand-chosen design choices and domain assumptions are listed above. The key assumptions are that HiResCAM maps are faithful, that validation accuracy captures the regularization effect, and that the two datasets represent their domains.

free parameters (2)
  • candidate window percentile = 10%
    Low-saliency selects from the lowest-scoring 10% of valid windows and high-saliency from the highest-scoring 10%; this threshold is chosen by hand, is not swept, and controls the diversity of cutout placements (Section 3.3).
  • validation split fraction = 0.1
    Each run uses a 10% validation split; this choice affects the reported accuracy estimates and is a hand-selected experimental setting (Section 3.6).
assumptions (4)
  • domain assumption HiResCAM maps are faithful indicators of class evidence.
    The method selects cutouts based on HiResCAM saliency; if the maps do not reflect task-relevant regions, the comparison tests nothing about saliency-guided placement (Sections 2.2 and 3.3).
  • domain assumption Validation accuracy is an adequate proxy for the regularization effect.
    The paper reports only validation metrics and no held-out test set, and it explicitly notes this limitation (Sections 3.6 and 5.5). This assumption underlies all conclusions about improvement or degradation.
  • domain assumption RawMal-TF grayscale images and CIFAR-100 are representative of their domains.
    The domain-dependence conclusion is drawn from one malware dataset and one natural-image benchmark; the paper restricts its claims accordingly, but generalization from these two datasets remains an assumption (Sections 3.5 and 5.5).
  • domain assumption The no-cutout teacher checkpoint is a valid source of saliency for student cutout placement.
    CAM-guided cutout relies on saliency maps from a teacher trained without cutout; the paper does not test whether another teacher source would change the conclusions (Sections 3.2 and 5.5).

how reviews work

0 comments
Cite this review

Pith. "Pith review of CAM-Guided Saliency Cutout and Image-Based Malware Classification." pith.science (2026). https://pith.science/paper/GDDPJJLG

@misc{pith2026260811634,
  author       = {Pith},
  title        = {Pith review of: CAM-Guided Saliency Cutout and Image-Based Malware Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GDDPJJLG}},
  note         = {Machine review of arXiv:2608.11634}
}
abstract

Dropout regularization is commonly used to reduce overfitting by removing parts of a neural network during training. For Convolutional Neural Networks (CNN), cutouts serve a somewhat analogous purpose. Cutouts can be implemented as data augmentation: the original training image is retained, and additional copies are created with regions removed. In this chapter, we test whether cutout placement can be improved by using High-Resolution Class Activation Mapping (HiResCAM). We compare four controlled training conditions: no cutout, standard random cutout, low-saliency cutout, and high-saliency cutout. We experiment using grayscale malware images from the RawMal-TF dataset (17 families with~1,000 samples per family), and for comparison to natural images, we experiment with the well-known CIFAR-100 dataset. All experiments are based on ResNet18 with~100 training epochs. For the cutout experiments, we test cutout areas of~5\%, 10\%, 20\%, and~30\%, and we consider~$M\in\{4,8}$ augmented copies per original training image. The RawMal-TF results are slightly worse for all three cutout cases (random, high and low saliency) as compared to no cutouts. In contrast, our CIFAR-100 experimental results improve slightly under low-saliency cutout. These results suggest that the value of saliency-guided cutout is domain dependent, and that malware images should not be treated as equivalent to natural images.

Figures

Figures reproduced from arXiv: 2608.11634 by the authors.

Figure 1
Figure 1. Controlled cutout pipeline 3.3 CAM-Guided Cutout Implementation The saliency-guided cutout implementation is designed to mimic standard cutout as closely as possible. For a given image, cutout size, and augmentation index, the method selects one square window and fills it with normalized (i.e., black) value. The random condition selects this window randomly, while the CAM-guided conditions select it using a HiResCAM… view at source ↗
Figure 2
Figure 2. RawMal-TF best validation error by cutout area (lower is better; error [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. RawMal-TF paired effects relative to matched random cutout (positive [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: RawMal-TF paired-effect heatmap (blue cells favor saliency-guided [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: CIFAR-100 best validation error by cutout area (lower is better; error [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: CIFAR-100 paired effects relative to matched random cutout (positive [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: CIFAR-100 paired-effect heatmap (blue cells favor saliency-guided [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]
Figure 8
Figure 8. Figure 8: Seed-level paired effects for low-saliency cutout relative to matched [PITH_FULL_IMAGE:figures/full_fig_p015_8.png]
Figure 9
Figure 9. Figure 9: Best mean validation accuracy within each method family. Each cutout [PITH_FULL_IMAGE:figures/full_fig_p016_9.png]
Figure 10
Figure 10. Figure 10: Secondary learning-curve metrics for the selected method-family rep [PITH_FULL_IMAGE:figures/full_fig_p016_10.png]
Figure 11
Figure 11. Figure 11: Representative individual-run validation error curves. Aggregate con [PITH_FULL_IMAGE:figures/full_fig_p017_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 14 canonical work pages

  1. [1]

    A comparison of selected image transformation techniques for malware classification

    Rishit Agrawal, Kunal Bhatnagar, Andrew Do, Ronnit Rana, Martin Jureˇ cek, and Mark Stamp. A comparison of selected image transformation techniques for malware classification. In Roberto Di Pietro, Karen Renaud, and Paolo Mori, editors,Proceedings of the 12th International Conference on Information Systems Security and Privacy, volume 2 ofICISSP, pages 33...

    2026

  2. [2]

    A Comparison of Selected Image Transformation Techniques for Malware Classification

    Rishit Agrawal, Kunal Bhatnagar, Andrew Do, Ronnit Rana, and Mark Stamp. A comparison of selected image transformation techniques for malware classi- fication.https://arxiv.org/abs/2509.10838, 2025

    work page Pith review arXiv 2025

  3. [3]

    RawMal-TF: Raw malware dataset labeled by type and family.https://arxiv.org/abs/2506.23909, 2025

    David B´ alik, Martin Jureˇ cek, and Mark Stamp. RawMal-TF: Raw malware dataset labeled by type and family.https://arxiv.org/abs/2506.23909, 2025

    arXiv 2025

  4. [4]

    Trans- fer learning for image-based malware classification

    Niket Bhodia, Pratikkumar Prajapati, Fabio Di Troia, and Mark Stamp. Trans- fer learning for image-based malware classification. In Paolo Mori, Steven Fur- nell, and Olivier Camp, editors,Proceedings of the 5th International Conference on Information Systems Security and Privacy, ICISSP, pages 719–726, 2019

    2019

  5. [5]

    Through the static: Demystifying malware visualization via explainability.Journal of Information Security and Applications, 91:104063, 2025

    Matteo Brosolo, Vinod P., and Mauro Conti. Through the static: Demystifying malware visualization via explainability.Journal of Information Security and Applications, 91:104063, 2025

    work page 2025

  6. [6]

    Terrance DeVries and Graham W. Taylor. Improved regularization of convo- lutional neural networks with cutout.https://arxiv.org/abs/1708.04552, 2017

    arXiv 2017

  7. [7]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition at scale. InInterna- tional Conference on Learning Representations, ICLR, 2021

    work page 2021

  8. [8]

    Use HiResCAM instead of Grad- CAM for faithful explanations of convolutional neural networks.https:// arxiv.org/abs/2011.08891, 2020

    Rachel Lea Draelos and Lawrence Carin. Use HiResCAM instead of Grad- CAM for faithful explanations of convolutional neural networks.https:// arxiv.org/abs/2011.08891, 2020. 21

    arXiv 2011

Show all 26 references
  1. [9]

    CAMRegularization, GitHub repository.https:// github.com/yasamanebrahimi-byte/CAMRegularization, 2026

    Yasaman Ebrahimi. CAMRegularization, GitHub repository.https:// github.com/yasamanebrahimi-byte/CAMRegularization, 2026

    2026

  2. [10]

    Computing and evaluating saliency maps for image classification: a tutorial.Journal of Electronic Imaging, 32(2):020801, 2023

    Tristan Gomez and Harold Mouch` ere. Computing and evaluating saliency maps for image classification: a tutorial.Journal of Electronic Imaging, 32(2):020801, 2023

    2023

  3. [11]

    Naofumi Hama, Masayoshi Mase, and Art B. Owen. Deletion and insertion tests in regression models.Journal of Machine Learning Research, 24:1–38, 2023

    2023

  4. [12]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. InProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, CVPR, pages 770–778, 2016

    2016

  5. [13]

    Weinberger

    Gao Huang, Zhuang Liu, Laurens Van Der Maaten, and Kilian Q. Weinberger. Densely connected convolutional networks. InProceedings of the IEEE Confer- ence on Computer Vision and Pattern Recognition, CVPR, pages 4700–4708, 2017

    2017

  6. [14]

    Andreopoulos, and Mark Stamp

    Mugdha Jain, William B. Andreopoulos, and Mark Stamp. Convolutional neu- ral networks and extreme learning machines for malware classification.Journal of Computer Virology and Hacking Techniques, 16(3):229–244, 2020

    2020

  7. [15]

    Learning multiple layers of features from tiny images

    Alex Krizhevsky. Learning multiple layers of features from tiny images. Tech- nical report, University of Toronto, 2009

    2009

  8. [16]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. InProceedings of the IEEE/CVF International Conference on Computer Vision, ICCV, pages 10012–10022, 2021

    2021

  9. [17]

    A convnet for the 2020s

    Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Dar- rell, and Saining Xie. A convnet for the 2020s. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR, pages 11976– 11986, 2022

    2022

  10. [18]

    Kimmell, Mahmoud Ab- delsalam, Sajad Khorsandroo, and Maanak Gupta

    Harikha Manthena, Shaghayegh Shajarian, Jeffrey C. Kimmell, Mahmoud Ab- delsalam, Sajad Khorsandroo, and Maanak Gupta. Explainable artificial intel- ligence (XAI) for malware analysis: A survey of techniques, applications, and open challenges.IEEE Access, 13:61611–61640, 2025

    2025

  11. [19]

    Malware images: Visualization and automatic classification

    Lakshmanan Nataraj, Shankarapani Karthikeyan, George Jacob, and BS Man- junath. Malware images: Visualization and automatic classification. InPro- ceedings of the 8th International Symposium on Visualization for Cyber Secu- rity, pages 1–7, 2011

    2011

  12. [20]

    Advancing mal- ware imagery classification with explainable deep learning: A state-of-the-art approach using SHAP, LIME and Grad-CAM.PLOS ONE, 20(5):e0318542, 2025

    Sadia Nazim, Muhammad Mansoor Alam, Syed Safdar Rizvi, Jawahir Che Mustapha, Syed Shujaa Hussain, and Mazliham Mohd Suud. Advancing mal- ware imagery classification with explainable deep learning: A state-of-the-art approach using SHAP, LIME and Grad-CAM.PLOS ONE, 20(5):e03185...

    2025

  13. [21]

    RISE: Randomized input sampling for explanation of black-box models

    Vitali Petsiuk, Abir Das, and Kate Saenko. RISE: Randomized input sampling for explanation of black-box models. InBritish Machine Vision Conference, BMVC, page 151, 2018

    2018

  14. [22]

    Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, and Dhruv Batra

    Ramprasaath R. Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, and Dhruv Batra. Grad-CAM: Visual explanations from deep networks via gradient-based localization. In2017 IEEE International Conference on Computer Vision, ICCV, pages 618–626, 2017

    2017

  15. [23]

    Hinton, Alex Krizhevsky, Ilya Sutskever, and Ruslan Salakhutdinov

    Nitish Srivastava, Geoffrey E. Hinton, Alex Krizhevsky, Ilya Sutskever, and Ruslan Salakhutdinov. Dropout: A simple way to prevent neural networks from overfitting.Journal of Machine Learning Research, 15:1929–1958, 2014

    1929

  16. [24]

    A selective survey of deep learning techniques and their ap- plication to malware analysis

    Mark Stamp. A selective survey of deep learning techniques and their ap- plication to malware analysis. In Mark Stamp, Mamoun Alazab, and Andrii Shalaginov, editors,Malware Analysis Using Artificial Intelligence and Deep Learning, pages 3–51. Springer, 2021

    2021

  17. [25]

    Efficientnet: Rethinking model scaling for con- volutional neural networks

    Mingxing Tan and Quoc Le. Efficientnet: Rethinking model scaling for con- volutional neural networks. InInternational Conference on Machine Learning, ICML, pages 6105–6114, 2019

    2019

  18. [26]

    Deep learning versus gist descriptors for image-based malware classifi- cation

    Sravani Yajamanam, Vikash Raja Samuel Selvin, Fabio Di Troia, and Mark Stamp. Deep learning versus gist descriptors for image-based malware classifi- cation. In Paolo Mori, Steven Furnell, and Olivier Camp, editors,Proceedings of the 4th International Conference on Information...

    2018

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.