Giter Club home page Giter Club logo

sml20's People

Contributors

tbohne avatar

Watchers

 avatar

sml20's Issues

Review 1 Suggestions

  • could improve the first and second chapter of this paper
  • the introduction is very short and doesn’t provide enough information to make it understandable what the main problem and motivation is
  • a part of the text of the second chapter, especially the complete “motivation” section of chapter 2 could be integrated into the introduction
  • in the second section it could be explained more clearly how GNNs work and what they do
  • currently the abstract of this paper is missing and should be added
  • the list of challenges in 4.2 has the label of a figure but is no figure

Review 3 Suggestions

Es wäre gut, insbesondere folgende Aspekte noch etwas näher zu betrachten:

  • Abstract fehlt, Introduction ist recht kurz - hier ggf. über Überblick/Motivation nachdenken
  • Bei den "Paragraph-Umgebungen" (\paragraph{}) ist es durchaus erlaubt, dass der nachfolgende Text in der gleichen Zeile beginnt. Gut ist hier beispielsweise folgendes: \paragraph{XYZ-Paragraph.} asdf
  • "Figure 8: Major challenges to be solved." ???
  • Lassen sich die Conclusions noch stärker in gegliederte Textabschnitte teilen (Übersichtlichkeit)?
  • Literaturangaben - sind diese vollständig?
    • teilweise Teile von Quellenangaben fehlten, bitte prüfen Sie hier noch einmal sorgfältig, ob hier alles vollständig ist

Title suggestion

"Robustness Certification for Graph Structure Perturbations"

instead of

"Robustness of GCNs with respect to perturbations of the graph structure"

--> for the other one as well

Struktur Feedback

Feedback:
"Bitte ggf. über die Struktur/roter Faden in den Abschnitten 3 & 4 nachdenken, und diese ggf. "stärken"."

Refine literature references

  • check for each paper whether there is an actual journal reference (besides arxiv)
  • check style / order / names etc.

Introduction Feedback

Feedback:
"gute Übersicht, allerdings genauer, was hier noch alles kommt - und (teilweise) auch wie es gemacht wird."

Add to literature review section

To the best of their knowledge, the Zügner 2019 paper, the Bojchevski and the Zügner 2020 paper are the only existing works to study certified robustness of GNNs for node classification.

  • Wang 2020 Paper

Check the whole paper for undeclared citations (direct or indirect)

Feedback:
"Ich werde auch auf Plagiate im Text prüfen. Bei wörtlichen Übernahmen bitte genau zitieren. Bei Ihrer Version gibt es momentan sehr viele Textstellen, die fast wortwörtlich aus den entsprechenden Papers übernommen sind - ohne Referenz bzw. ohne ganz explizite Kennzeichnung; bei wörtlichen Übernahmen ist auch eine explizite Kennzeichnung erforderlich - alles andere wäre nicht zulässig. Im folgenden ein paar Beispiele, bei denen manchmal nur ein Wort im Vergleich zum Original-Text abweicht:"

  • these tasks require dealing with non-Euclidean graph data that contains rich relational information between elements and cannot be well handled by traditional deep learning models
  • By adding an imperceptibly small vector whose elements are equal to the sign of the elements of
    the gradient of the cost function with respect to the input, they can change GoogleNet’s classification of the image
  • Moreover, to remedy the propagation of adversarial attacks in GCNs, a variance-based attention mechanism is proposed, i.e. assigning different weights to node neighborhoods according to their variances when performing convolutions.
  • If a node has been certified with the method, it is guaranteed to be robust under any possible perturbation given the attack model.

Introduce GCNs

  • briefly explain what GCNs are and how they specialize general GNNs

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.