Pith. sign in

REVIEW 4 major objections 6 minor 33 references

A Graph Based Raman Spectral Processing Technique for Exosome Classification

T0 review · 4 major / 6 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 A graph-based PageRank filter plus forward feature selection lifts Raman exosome classification accuracy to 0.760 on spectra and 0.857 on surfaces under group cross-validation.

desk verdict Plausible new Raman preprocessing combination, but the reported accuracy gains may be inflated because key tuning decisions are not demonstrably nested inside the cross-validation. read the letter →

arxiv 2504.15324 v1 pith:CH7QTUUS submitted 2025-04-21 q-bio.QM cs.AIcs.ITcs.LGmath.IT

Vuong M. Ngo , Edward Bolger , Stan Goodwin , John O'Sullivan , Dinh Viet Cuong , Mark Roantree This is my paper · ORCID
classification q-bio.QMcs.AIcs.ITcs.LGmath.IT
keywords ExosomeclassificationRamanspectroscopySurface-enhancedPageRankfilterGraphdatabaseSpectralpreprocessingForwardfeatureselectionGroupcross-validation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper claims that standard Raman preprocessing leaves noisy, unrepresentative spectra in exosome datasets, and that this noise can be removed with a graph-based filter: each spectrum is placed in a graph built from Raman peaks, scored by PageRank centrality, and low-scoring spectra are discarded. Adding a forward sequential feature selection that keeps the 50 most informative wave numbers then gives a compact feature set that classifies exosomes from normal, hyperglycemic, and hypoglycemic cells at 0.760 accuracy on Raman spectra and 0.857 accuracy on Raman surfaces under group 10-fold cross-validation. The practical interest is that exosomes circulate in blood, saliva, and urine, so a reliable Raman-plus-graph pipeline would support liquid-biopsy diagnostics without invasive tissue sampling. For the reported numbers to be unbiased, every tuning choice in the pipeline must be made inside each training fold.

What carries the argument

The load-bearing mechanism is the PageRank Filter (PRF) over a peak-based graph. Each Raman peak is a node; edges connect peaks that come from the same spectrum or that lie close in wave number, so the graph encodes co-occurrence and spectral proximity. PageRank centrality—a score that ranks nodes by their connectedness in the graph—is computed for each spectrum, and spectra below a centrality cutoff are dropped as noise. The dimensionality-reduction stage then runs a forward sequential search over wave numbers, keeping the 50 that most improve model performance. Extra Trees is the classifier that best exploits this reduced representation in the paper's experiments.

What would settle it

Repeat the full evaluation with nested group 10-fold cross-validation, re-selecting the interquartile-range cutoff, the PageRank threshold, the Gaussian kernel width, and the 50 wave numbers inside each training fold and applying them only to the held-out group. If held-out accuracy drops to the OSC-only baseline (0.635 spectra, 0.683 surfaces), the reported gains are an artifact of tuning on the evaluation groups.

Watch

Extended reading notes

Core claim

The authors' central discovery is that a three-stage pipeline—standard cleaning (cosmic-spike removal, baseline correction, smoothing, scaling), a PageRank Filter that removes spectra with low graph centrality, and an optimal dimensionality-reduction step that forward-selects 50 wave numbers—classifies exosome samples substantially better than standard cleaning alone. On the raw Raman spectra, the pipeline reaches 0.760 accuracy with an Extra Trees model versus 0.635 for cleaning alone; on the SERS surfaces, it reaches 0.857 versus 0.683. The authors attribute the gain to the graph filter suppressing unrepresentative spectra while the feature selection keeps biomarker-relevant peaks, and they report that surface-level Raman data are consistently easier to classify than single-spectrum data.

Load-bearing premise

The reported accuracies assume that every tuning decision—the interquartile-range outlier cutoff, the PageRank centrality threshold, the Gaussian kernel width, and the 50 selected wave numbers—was made inside each training fold of the group 10-fold cross-validation, so that no held-out information influenced the final model.

Editorial extensions

If this is right

  • The complete OSC+PRF+DR pipeline improves accuracy from 0.635 to 0.760 on spectra and from 0.683 to 0.857 on surfaces under group 10-fold cross-validation.
  • The PageRank filter alone accounts for part of the gain (0.696 on spectra, 0.825 on surfaces), so graph-based outlier removal helps even before feature selection is added.
  • Surface-level Raman data classify more cleanly than single spectra under every preprocessing variant, including perfect recall for the hyperglycemic class and perfect precision for the normal class under the full pipeline.
  • Because group 10-fold cross-validation never splits spectra from the same surface across training and test sets, the reported accuracies are estimates for classifying previously unseen experimental surfaces rather than new spectra from known surfaces.

Reading between the lines

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

  • The decisive test the paper leaves undone is nested group cross-validation: if the interquartile-range cutoff, PageRank threshold, kernel width, and 50 selected wave numbers were tuned on held-out groups, reported accuracies would be optimistic; a nested re-run would settle whether the gains persist.
  • The graph encodes more information than the filter uses: node embeddings or centrality features could be fed directly to the classifier, so the paper's pipeline likely under-exploits the graph structure it builds.
  • The 50 selected wave numbers are a testable biomarker claim: checking them against known exosome Raman bands (amide I/III, CH2, phosphate backbone) would show whether the pipeline preserves molecular signal or merely separates surface preparation batches.
  • The gap between surface-level and spectrum-level accuracy suggests the SERS surface itself carries predictive information; distinguishing a true exosome signal from a batch effect would require surfaces prepared from independent biological replicates.
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

4 major / 6 minor

Summary. The paper proposes a three-stage Raman spectral processing pipeline, OSC+PRF+DR, for classifying exosomes from normal, hyperglycemic, and hypoglycemic cells. OSC is a standard cleaning sequence (despiking, baseline correction, smoothing, scaling); PRF adds graph-based PageRank filtering of spectra per SERS surface, after constructing Gaussian-kernel similarity graphs in Neo4j; DR applies forward sequential feature selection to retain 50 wavenumbers. The authors evaluate with group 10-fold cross-validation on 3,045 spectra from 63 SERS surfaces, using SVM, Random Forest, and Extra Trees. They report accuracies of 0.760 for spectra and 0.857 for surfaces with OSC+PRF+DR, compared with 0.635 and 0.683 for the OSC baseline, and claim that graph-based filtering plus dimensionality reduction significantly improves classification.

Significance. If the reported gains are real, the combination of graph-centrality-based outlier removal and feature selection would be a useful addition to Raman spectral preprocessing, and the use of group cross-validation is methodologically commendable. The paper also introduces a publicly relevant biomedical task and presents a clearly structured comparison of three processing pipelines. The main weakness is that the evaluation protocol does not demonstrate that all tuning decisions are nested inside the cross-validation folds; the paper's own conclusion lists nested cross-validation as future work. Because the reported accuracy improvements are the paper's central claim, this is a load-bearing issue. The absence of any variance estimates or repeated cross-validation further weakens the claim that the differences are significant. The core idea is plausible, but the current evidence does not yet establish superiority over the OSC baseline.

major comments (4)
  1. [Section 3.3, 3.4, 5] The tuning decisions—the IQR proportion cutoff, the PageRank centrality threshold, the Gaussian kernel width, and the number of selected wavenumbers—are not stated to be chosen inside each training fold of the group 10-fold cross-validation. The text in Section 3.3 says the IQR proportion cutoff was 'systematically tested' and a PageRank cutoff was applied, and Section 3.4 says features were selected by forward sequential selection 'iteratively selecting the wave numbers that contributed most to model performance.' The conclusion explicitly lists nested cross-validation as future work, which is an in-text admission that the current evaluation is not nested. If any of these choices used held-out groups, the reported accuracies are optimistically biased, and the comparison against OSC (which receives no comparable tuning) is unfair. Please either demonstrate that every tuning step was performed inside the training folds, or re-run the evaluation with fully nested cross-validation and report the resulting accuracies.
  2. [Tables 1 and 2] The reported accuracies are point estimates with no standard deviations, confidence intervals, or repeated cross-validation runs. With only 63 unique SERS surfaces split into 10 groups, the observed differences (0.760 vs 0.635 for spectra; 0.857 vs 0.683 for surfaces) may be within sampling noise. The conclusion describes OSC+PRF+DR as 'significantly better' and 'significantly improved,' but no statistical test or variance estimate supports this wording. Please report per-fold results, repeated group-10-fold cross-validation with mean and standard deviation, and, ideally, a paired significance test across folds (or another appropriate procedure).
  3. [Section 3.3] The PageRank filtering procedure is underspecified. It is unclear whether the IQR proportion cutoff and the PageRank centrality cutoff are global constants applied identically to every surface, or are adaptively chosen per surface. The Gaussian kernel width used to build the similarity subgraphs is never reported, although it determines which spectra are considered close enough to share an edge. The choice of the cutoff values affects how many spectra survive filtering and, consequently, all downstream accuracy numbers. Please give the exact parameter values used in the final experiments, the range tested, and the criterion used to pick the final values, and explain whether these choices were made per surface or globally.
  4. [Section 3.4] The forward sequential feature selection is not described precisely enough to assess whether it leaks information. The text says wave numbers are selected 'iteratively' based on 'model performance,' but it does not state what performance measure is used, what inner validation procedure (if any) is employed, or how the 50-feature limit interacts with the group cross-validation. If the feature selection is run on the entire dataset before the outer cross-validation loop, the chosen wavenumbers carry information from the held-out groups, biasing the reported accuracy. Feature selection must be embedded inside each training fold (or in a nested inner loop), and the selected wavenumbers should be reported for at least one fold to assess stability.
minor comments (6)
  1. [Figure 5 caption] The caption contains a typo: 'Extra Tress model' should read 'Extra Trees model.'
  2. [Figure 2 caption] The caption says 'after OSC proccesing'; 'proccesing' should be 'processing.'
  3. [Section 3.2] The sentence 'To minimize high-frequency noise, the Savitzky-Golay filter, a robust digital signal processing technique widely used in spectral analysis to preserve signal characteristics while reducing noise [9].' is grammatically incomplete; a main verb is missing.
  4. [Section 3.4 and Tables 1-2] The 50 selected wavenumbers are never listed or characterized, making it difficult to judge biological plausibility or reproducibility. Please provide the selected wavenumber indices or a representative list, and discuss whether the selection is stable across folds.
  5. [Section 4.2] The 'Best Model' row of Tables 1 and 2 reports only the best-performing classifier, but the manuscript does not give the hyperparameters used for SVM, Random Forest, or Extra Trees. Without this information, the comparison across pipelines cannot be fully reproduced.
  6. [Section 3.1] The benefit of using Neo4j as a graph database, as opposed to constructing the same graphs in memory, is not explained. If the graph database is incidental to the method, please clarify this; otherwise, describe what graph features or computational advantages it provides.

Circularity Check

1 steps flagged · score 4.0 of 10

Reported OSC+PRF+DR accuracy gains are partly self-evaluative: the forward feature selector optimizes the same performance metric later reported, and nested cross-validation is deferred to future work. No derivation-level circularity found.

  1. fitted input called prediction [Section 3.4 (feature selection); Section 3.3 (IQR/PageRank cutoffs); Section 4.2 (Tables 1-2); Section 5 (future work).]
    "employing a forward sequential feature selection process, iteratively selecting the wave numbers that contributed most to model performance."

    The 50 wavenumbers are chosen by optimizing model performance; the same metric is then reported in Tables 1 and 2 as group-10-fold validation. Unless the selection, the IQR proportion cutoffs that were 'systematically tested' for improved performance (Section 3.3), and the PageRank centrality cutoff were all re-estimated inside each training fold, the held-out groups' labels have influenced the feature set, making the reported accuracy a fit to the evaluation data rather than an independent prediction. The conclusion's statement that 'using nested cross-validation would further validate the OSC+PRF and OSC+PRF+DR' is an in-text admission that this model-selection loop is not currently nested.

full rationale

No derivation-level circularity: no equation reduces to its input, no self-citation chain is load-bearing, and no ansatz is presented as an external theorem. The graph construction, PageRank centrality, and Extra Trees classification are genuine data-processing choices with independent content. The one substantive circular-evaluation burden is the target-based tuning of the feature subset and outlier/centrality cutoffs, which is not demonstrated to be nested inside the group-10-fold cross-validation and is explicitly deferred to future work as nested cross-validation. This makes the headline accuracy numbers partly self-evaluative, although the central methodological claim that graph-based spectral filtering can aid classification remains falsifiable and testable with a properly nested protocol. The score is 4 rather than 0 because the tuning-evaluation overlap is a real, quoted reduction of the reported metric to the selection objective; it is not higher because the pipeline itself is not derived from the results.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

The method has no mathematical derivation; its performance rests on several tuned parameters and domain assumptions. The core contribution is empirical, so the ledger lists the parameters chosen against the data and the domain beliefs that the approach relies on.

free parameters (5)
  • IQR outlier proportion cutoff = not reported (systematically tested)
    Spectra are discarded if a proportion of values fall outside 1.5*IQR; the proportion cutoff was tuned on the dataset (Section 3.3).
  • PageRank centrality threshold = not reported
    Spectra below a centrality threshold are removed; threshold is chosen by model performance and not reported (Section 3.3).
  • Gaussian kernel width = not reported
    Constructs per-surface spectral similarity graph; scale parameter unspecified (Section 3.3).
  • Number of selected wave numbers = 50
    Forward sequential feature selection capped at 50 features for computational efficiency; feature subset selected by model performance (Section 3.4).
  • Peak proximity threshold = not reported
    Edges between peaks are added if peaks are 'close in proximity'; threshold unspecified (Section 3.1).
assumptions (3)
  • domain assumption Raman spectral peaks correspond to biomarker identity and proximity in wave number implies potential shared biomarker
    Graph construction in Section 3.1 links peaks close in proximity on the assumption they correspond to the same biomarker.
  • domain assumption PageRank centrality on a Gaussian kernel spectral graph ranks spectra by representativeness of the underlying Raman signal
    Section 3.3 assumes higher PageRank means better representative spectra and low PageRank means noise.
  • domain assumption Group 10-fold cross-validation with surfaces as groups yields unbiased generalization estimates for the reported accuracies
    Section 4.1 adopts this protocol, but it only prevents leakage if all tuning happens inside the folds, which is not established.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Graph Based Raman Spectral Processing Technique for Exosome Classification." pith.science (2026). https://pith.science/paper/CH7QTUUS

@misc{pith2026250415324,
  author       = {Pith},
  title        = {Pith review of: A Graph Based Raman Spectral Processing Technique for Exosome Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CH7QTUUS}},
  note         = {Machine review of arXiv:2504.15324}
}
read the original abstract

Exosomes are small vesicles crucial for cell signaling and disease biomarkers. Due to their complexity, an "omics" approach is preferable to individual biomarkers. While Raman spectroscopy is effective for exosome analysis, it requires high sample concentrations and has limited sensitivity to lipids and proteins. Surface-enhanced Raman spectroscopy helps overcome these challenges. In this study, we leverage Neo4j graph databases to organize 3,045 Raman spectra of exosomes, enhancing data generalization. To further refine spectral analysis, we introduce a novel spectral filtering process that integrates the PageRank Filter with optimal Dimensionality Reduction. This method improves feature selection, resulting in superior classification performance. Specifically, the Extra Trees model, using our spectral processing approach, achieves 0.76 and 0.857 accuracy in classifying hyperglycemic, hypoglycemic, and normal exosome samples based on Raman spectra and surface, respectively, with group 10-fold cross-validation. Our results show that graph-based spectral filtering combined with optimal dimensionality reduction significantly improves classification accuracy by reducing noise while preserving key biomarker signals. This novel framework enhances Raman-based exosome analysis, expanding its potential for biomedical applications, disease diagnostics, and biomarker discovery.

Figures

Figures reproduced from arXiv: 2504.15324 by the authors.

Figure 1
Figure 1. The Emission of Exosomes from a Cell [27] Due to their small size (40–150 nm) and presence in complex biological sam￾ples, exosomes require specialized tech￾niques for measurement and analysis [17]. Raman Spectroscopy, a non-destructive vibrational spectroscopic technique based on inelastic photon scattering, is par￾ticularly effective for analyzing exosome structures [7]. Using a monochromatic laser, it excites che… view at source ↗
Figure 2
Figure 2. Examples of Raman spectra after OSC proccesing [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. The 10% most and least central spectra within a surface according to [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Average spectra by status via OSC+PRF+DR processing [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Average performance per class of OSC+PRF+DR approach [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 33 canonical work pages

  1. [1]

    PeerJ Computer Science9, e1333 (2023)

    Alqaissi, E., Alotaibi, F., Ramzan, M.S.: Graph data science and machine learning for the detection of covid-19 infection from symptoms. PeerJ Computer Science9, e1333 (2023)

    work page 2023

  2. [2]

    Vibrational Spectroscopy99, 124–129 (2018)

    Amjad, A., Ullah, R., Khan, S., Bilal, M., Khan, A.: Raman spectroscopy based analysis of milk using random forest classification. Vibrational Spectroscopy99, 124–129 (2018)

    work page 2018

  3. [3]

    Computer Networks 56(18), 3825–3833 (2012)

    Brin, S., Page, L.: The anatomy of a large-scale hypertextual web search engine. Computer Networks 56(18), 3825–3833 (2012)

    work page 2012

  4. [4]

    In: CIKM’19

    Chen, H., Sultan, S.F., Tian, Y., Chen, M., Skiena, S.: Fast and accurate network embeddings via very sparse random projection. In: CIKM’19. p. 399–408 (2019)

    work page 2019

  5. [5]

    Scientific Re- ports 13(1) (2023)

    Chen, X., et al.: Raman spectroscopy combined with a support vector machine algorithm as a diagnostic technique for primary sjögren’s syndrome. Scientific Re- ports 13(1) (2023)

    work page 2023

  6. [6]

    IEEE Open Journal of Intelligent Transportation Systems 5, 115–131 (2024)

    Cuong, D.V., Ngo, V.M., Cappellari, P., Roantree, M.: Analyzing shared bike usage through graph-based spatio-temporal modeling. IEEE Open Journal of Intelligent Transportation Systems 5, 115–131 (2024)

    work page 2024

  7. [7]

    Vibrational Spectroscopy57(2), 163–176 (2011)

    Das, R.S., Agrawal, Y.K.: Raman spectroscopy: Recent advancements, techniques and applications. Vibrational Spectroscopy57(2), 163–176 (2011)

    work page 2011

  8. [8]

    Eilers, P., Boelens, H.: Baseline correction with asymmetric least squares smooth- ing (2005)

    work page 2005

Show all 33 references
  1. [9]

    Gallagher, N.: Savitzky-golay smoothing and differentiation filter (2020)

    2020

  2. [10]

    EPJ Techniques and Instrumentation 2(1) (2015)

    Gautam, R., Vanga, S., Ariese, F., Umapathy, S.: Review of multidimensional data processing approaches for raman and infrared spectroscopy. EPJ Techniques and Instrumentation 2(1) (2015)

    2015

  3. [11]

    SIAM Review57(3), 321–363 (2015)

    Gleich, D.F.: Pagerank beyond the web. SIAM Review57(3), 321–363 (2015)

    2015

  4. [12]

    Analytical Methods9(30) (2017)

    Guo, S., Bocklitz, T., Neugebauer, U., Popp, J.: Common mistakes in cross- validating classification models. Analytical Methods9(30) (2017)

    2017

  5. [13]

    Scientific Reports12, 3456 (2022)

    Hu, S., et al.: Raman spectroscopy combined with machine learning algorithms to detect adulterated suichang native honey. Scientific Reports12, 3456 (2022)

    2022

  6. [14]

    Journal of Clinical Investigation 126(4), 1208–1215 (2016)

    Kalluri, R.: The biology and function of exosomes in cancer. Journal of Clinical Investigation 126(4), 1208–1215 (2016)

    2016

  7. [15]

    Science 367(6478) (2020)

    Kalluri, R., LeBleu, V.S.: The biology, function, and biomedical applications of exosomes. Science 367(6478) (2020)

    2020

  8. [16]

    Scientific Reports 12(1), 18561 (2022)

    Kopeć, M., Beton, K., Jarczewska, K., Abramczyk, H.: Hyperglycemia and cancer in human lung carcinoma by means of raman spectroscopy and imaging. Scientific Reports 12(1), 18561 (2022)

    2022

  9. [17]

    Acta Biomaterialia144, 1–14 (2022)

    Li, J., et al.: Exosome detection via surface-enhanced raman spectroscopy for can- cer diagnosis. Acta Biomaterialia144, 1–14 (2022)

    2022

  10. [18]

    Chemometrics and Intelligent Laboratory Systems 202, 104029 (2020)

    Martyna, A., et al.: Improving discrimination of raman spectra by optimising preprocessing strategies on the basis of the ability to refine the relationship be- tween variance components. Chemometrics and Intelligent Laboratory Systems 202, 104029 (2020)

    2020

  11. [19]

    Neo4j Inc.: Neo4j graph platform – the leader in graph databases (2023),https: //neo4j.com

    2023

  12. [20]

    Master’s thesis, Dublin City University, Ireland (2022)

    O’Sullivan, J.: Development of Bioplasmonic Platforms for Extracellular Vesicle Capture and Analysis. Master’s thesis, Dublin City University, Ireland (2022)

    2022

  13. [21]

    Qin, J., Xu, Q.: Functions and application of exosomes (2014)

    2014

  14. [22]

    Ruhnau, B.: Eigenvector-centrality — a node-centrality? Social Networks22(4), 357–365 (2000)

    2000

  15. [23]

    Journal of Raman Spectroscopy 45(10), 930–940 (2014)

    Schumacher, W., Stöckel, S., Rösch, P., Popp, J.: Improving chemometric results by optimizing the dimension reduction for raman spectral data sets. Journal of Raman Spectroscopy 45(10), 930–940 (2014)

    2014

  16. [24]

    ACS Nano14(5) (2020)

    Shin, H., et al.: Early-stage lung cancer diagnosis by deep learning-based spectro- scopic analysis of circulating exosomes. ACS Nano14(5) (2020)

    2020

  17. [25]

    Sensors20(5), 1317 (2020)

    Sierra, J., Marrugo-Ramírez, J., Rodriguez-Trujillo, R., Mir, M., Samitier, J.: Sensor-integrated microfluidic approaches for liquid biopsies applications in early detection of cancer. Sensors20(5), 1317 (2020)

    2020

  18. [26]

    Frontiers in Endocrinology 12 (2021)

    Sun, Y., Tao, Q., Wu, X., Zhang, L., Liu, Q., Wang, L.: The utility of exosomes in diagnosis and therapy of diabetes mellitus and associated complications. Frontiers in Endocrinology 12 (2021)

    2021

  19. [27]

    Frontiers in Genetics9 (2018)

    de la Torre Gomez, C., Goreham, R.V., Serra, J.J.B., Nann, T., Kussmann, M.: ‘exosomics’—a review of biophysics, biology and biochemistry of exosomes with a focus on human breast milk. Frontiers in Genetics9 (2018)

    2018

  20. [28]

    IEEE Transactions on Dielectrics and Electrical Insulation28(6), 1892–1900 (2021)

    Wang, Z., Chen, W., Zhou, W., Zhang, R., Song, R., Yang, D.: A few-shot learning method for aging diagnosis of oil-paper insulation by raman spectroscopy based on graph theory. IEEE Transactions on Dielectrics and Electrical Insulation28(6), 1892–1900 (2021)

    2021

  21. [29]

    Chemometrics and Intelligent Laboratory Systems179, 82–84 (2018)

    Whitaker, D.A., Hayes, K.: A simple algorithm for despiking raman spectra. Chemometrics and Intelligent Laboratory Systems179, 82–84 (2018)

    2018

  22. [30]

    Nano Letters 22(19), 7910–7918 (2022)

    Xie, Y., Su, X., Wen, Y., Zheng, C., Li, M.: Artificial intelligent label-free sers pro- filing of serum exosomes for breast cancer diagnosis and postoperative assessment. Nano Letters 22(19), 7910–7918 (2022)

    2022

  23. [31]

    Zhang, J., Luo, Y.: Degree centrality, betweenness centrality, and closeness central- ityinsocialnetwork.In:the2ndInternationalConferenceonModelling,Simulation and Applied Mathematics (MSAM 2017) (2017)

    2017

  24. [32]

    Spectrochimica Acta Part A: Molecular and Biomolecular Spec- troscopy 264, 120300 (2022)

    Zhang, L., et al.: Raman spectroscopy and machine learning for the classification of breast cancers. Spectrochimica Acta Part A: Molecular and Biomolecular Spec- troscopy 264, 120300 (2022)

    2022

  25. [33]

    ACM Computing Surveys55(2), 1–35 (2022)

    Zhou, J., Liu, L., Wei, W., Fan, J.: Network representation learning: From prepro- cessing, feature extraction to node embedding. ACM Computing Surveys55(2), 1–35 (2022)

    2022

Pith tools

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