Models of Random Graphs and Their Applications to the Web-Graph Analysis

Райгородский, А. М.

doi:10.1007/978-3-319-41718-9_5

Cited by 6 publications

(6 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The degree assortativity coefficient [34] estimates the tendency of vertices to connect with other vertices with similar features within the graph and was computed as the Pearson correlation coefficient between vertex degrees:…”

Section: Graph Invariants As Synthetic Featuresmentioning

confidence: 99%

Image Classification of Leukemic Cells Using Invariants of Triangle-Free Graphs as Synthetic Features

Muzalevskiy,

Torshin

2024

IEEE Access

View full text Add to dashboard Cite

Development of efficient methods of the cellular image processing is an important avenue for practical application of modern artificial intelligence techniques. In particular, practical hematology requires automatic classification of images with or without leukemic (blast) cells in peripheral blood smears. This paper presents a new approach to the problem of classification of such cellular images based on graph theory, XGBoost algorithm and convolutional neural networks (CNN). Firstly, each image is transformed into a weighted graph using gradient of intensity. Secondly, a number of graph invariants are computed thus producing a set of synthetic features that is used to train machine learning model based on XGBoost. Combining XGBoost with CNN further increases the accuracy of leukemic cell classification. Sensitivity (TPR) and Specifity (TNR) of the XGBoost-based model were 95% and 97% accordingly; ResNet-50 model showed TPR of 95% and TNR of 98%. Combined use of the XGBoost-based and the ResNet-50 models demonstrated TPR of 99% and TNR of 99%.

show abstract

Section: Graph Invariants As Synthetic Featuresmentioning

confidence: 99%

Image Classification of Leukemic Cells Using Invariants of Triangle-Free Graphs as Synthetic Features

Muzalevskiy,

Torshin

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

Section: Introductionmentioning

confidence: 99%

“…The object for these experiments can be either real or virtual social networks [4,5]. This article deals with virtual social networks, i.e., models of real social networks, and more precisely, the 7,8].Further, the article will be structured as follows: we will review similar works, consider a mathematical model of social…”

mentioning

confidence: 99%

How to Make the Information Dissemination Process More Effective: Studies based on Simulation Methods

Strekalovskaya,

Syrvacheva,

Zamyatina

2023

Proceedings of Computer Science and Information Technologies 2023 Conference

View full text Add to dashboard Cite

The article deals with issues related to the definition of parameters that affect the process of information dissemination in social media. Directed or bidirectional links are considered as parameters. The study of the propagation process is carried out using simulation methods. AnyLogic was chosen as a tool for conducting simulation experiments.Keywords-Social media, directed and not directed links, the degree of influence. I. INTRODUCTIONSocial media are web platforms and applications that allow to interact with the content of various types (text, photos, video and audio) and provide the user with the opportunity to interact with other users, rate and comment on content, and distribute it through likes, reposts and messages. In Russia, among the popular social media, the most popular is Vkontakte (VK). VK is a platform on which 9 out of 10 Russian users are registered. Another popular platform is the Odnoklassniki platform (7 out of 10 users). VK and Odnoklassniki allow users not only to communicate with each other, but also to comment on posts, organize events, etc., i.e. they are a channel for the dissemination of information, including the distribution of advertising products, services, materials for training, etc. [1]. The fast and reliable distribution of advertising is, without a doubt, beneficial for businesses. There is another side of this process -the dissemination of unwanted information. One way or another, the study of the process of dissemination of information, the parameters that influenced this process, or, conversely, the containment of information is relevant and important [2].Simulation modeling methods were chosen for studying and conducting research, and the AnyLogic[3] simulation modeling system was used as a tool. Unlike the methods used in SNA (Social Network Analyzes), which mainly study the structural characteristics of social media, simulation methods allow us to consider the process in dynamics, to identify the causes that can cause certain events. There are various approaches to conducting simulation experiments. The object for these experiments can be either real or virtual social networks [4,5]. This article deals with virtual social networks, i.e., models of real social networks, and more precisely, the 7,8].Further, the article will be structured as follows: we will review similar works, consider a mathematical model of social

show abstract

“…Этот процесс и стал основой первой части алгоритма, в которой генери-руются независимые сообщества с пользователями. Связи между пользователями образуются по модели Эрдеша-Реньи [4], как наиболее подходящей [3] для быстрого выполнения алгоритма.…”

Section: Introductionunclassified

Generation of a Social Network Graph by Using Apache Spark

Белов¹,

Вовчок²

2016

Model. anal. inf. sist.

View full text Add to dashboard Cite

Генерация графа социальной сети с использованием Apache SparkБелов Ю. А., Вовчок С.И. получена 24 октября 2016Аннотация. Планируется создать метод кластеризации графа социальной сети. Для тести-рования будущего метода возникла необходимость в генерации графа, по своей структуре схоже-го с лежащими в основе существующих социальных сетей. В статье представлен алгоритм для распределенной генерации такого графа. Учитываются основные свойства социальной сети: сте-пенное распределение количества сообществ для пользователей, плотные пересечения сообществ и другие. В данном алгоритме учтены проблемы, присутствующие в подобных работах других авторов, например, проблема кратных ребер при генерации. Особенностью созданного алгоритма стала реализация, зависящая от такого параметра как количество сообществ, а не от количе-ства пользователей, как это делается в других работах. Это связано с особенностью развития структуры реальной существующей социальной сети. В работе перечислены свойства ее графа. Описана таблица, содержащая необходимые для алгоритма переменные. Составлен пошаговый алгоритм генерации. Для него определены соответствующие математические параметры. Генера-ция происходит распределенно с помощью фреймворка Apache Spark. Подробно описано, каким образом происходит разделение задач с помощью данного фреймворка. В алгоритме использует-ся модель Эрдеша-Реньи для случайных графов как наиболее подходящая и достаточно простая для реализации. Основными преимуществами созданного метода являются использование малого количества ресурсов, по сравнению с другими подобными генераторами, и скорость выполнения. Быстрота достигается за счет распределенной работы и того, что при распределенной работе алго-ритма в любой момент пользователи сети имеют свои уникальные номера и упорядочены по этим номерам, поэтому не требуется их сортировка. Разработанный алгоритм будет способствовать не только созданию эффективного метода кластеризации. Он может быть полезен в других областях, связанных, например, с поисковыми системами социальных сетей.

show abstract

Models of Random Graphs and Their Applications to the Web-Graph Analysis

Cited by 6 publications

References 26 publications

Image Classification of Leukemic Cells Using Invariants of Triangle-Free Graphs as Synthetic Features

Image Classification of Leukemic Cells Using Invariants of Triangle-Free Graphs as Synthetic Features

How to Make the Information Dissemination Process More Effective: Studies based on Simulation Methods

Generation of a Social Network Graph by Using Apache Spark

Contact Info

Product

Resources

About