Special Issue "Applications of Big Data Analysis and Modeling"

A special issue of Mathematics (ISSN 2227-7390). This special issue belongs to the section "Probability and Statistics".

Deadline for manuscript submissions: 31 August 2023 | Viewed by 1022

Special Issue Editors

Computer Science Department, Southwest University, Tiansheng Road #2, Beibei District, Chongqing 400715, China
Interests: data-driven system modeling; network science; community detection; network representation learning; complex social networks analysis
Special Issues, Collections and Topics in MDPI journals
School of Artificial Intelligence, Optics and Electronics, Northwestern Polytechnical University, Xi’an 710072, China
Interests: intelligent decision-making and cognition; data mining; artificial intelligence; network science; complex systems
School of Artificial Intelligence, Optics and Electronics, Northwestern Polytechnical University, Xi’an 710072, China
Interests: data-driven modeling; social networks analysis; application of machine learning approaches
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

With the rapid development of modern internet technology, vast quantities of data are generated every day in practice, including texts, videos and images. For instance, the emergence and prosperity of various social platforms have witnessed the rapid development of online communications, especially information sharing.

In the era of big data, the core research fields of both academia and industry reside in data-driven applications, consisting of many machine learning approaches, such as supervised, semi-supervised and unsupervised learning, aiming to mine valuable information from the big data to provide convenience to our daily lives. Especially, graph analysis is becoming a hot topic that attracts various interests from biologists, economists, chemists, physicists, etc. The collected data in graph analysis can be represented by graph data through different embedding approaches, while many mathematics-based methods (e.g., matrix factorization) are developed. Recently, graph neural networks, which originate from spectral graph theory, generalize neural networks and deep learning to the graph. Due to the emergence of various deep learning models, the performance of analyzing data collected from practical systems can be efficiently promoted, such as recommendation, traffic forecasting, medicine development, epidemic spreading and natural language processing. The aim of this Special Issue is to publish cutting-edge original research papers on the latest advances in the analysis and application of big data in the development of machine learning approaches, including theories, models, algorithms, and applications in the real world.

Potential topics of interest include, but are not limited to:

  • Machine learning theories;
  • Machine learning models;
  • Machine learning algorithms;
  • Embedding/representation methods;
  • Feature selection and clustering;
  • Graph neural networks/graph convolutional networks;
  • Complex network analysis based on GNNs;
  • Sentiment analysis and text classification;
  • False account/news detection for online social networks;
  • Applications in NLP, emotional analysis, computer vision, intelligent traffic, recommendation system, financial, new medicine design, epidemiologic modeling, etc.

Prof. Dr. Chao Gao
Prof. Dr. Zhen Wang
Prof. Dr. Peican Zhu
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Mathematics is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2100 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • machine learning
  • deep learning
  • graph convolutional networks
  • statistical analysis of big data
  • applications of big data analysis
  • data-based complex network analysis

Published Papers (1 paper)

Order results
Result details
Select all
Export citation of selected articles as:

Research

Article
AdvSCOD: Bayesian-Based Out-Of-Distribution Detection via Curvature Sketching and Adversarial Sample Enrichment
Mathematics 2023, 11(3), 692; https://doi.org/10.3390/math11030692 - 29 Jan 2023
Viewed by 578
Abstract
Detecting out-of-distribution (OOD) samples is critical for the deployment of deep neural networks (DNN) in real-world scenarios. An appealing direction in which to conduct OOD detection is to measure the epistemic uncertainty in DNNs using the Bayesian model, since it is much more [...] Read more.
Detecting out-of-distribution (OOD) samples is critical for the deployment of deep neural networks (DNN) in real-world scenarios. An appealing direction in which to conduct OOD detection is to measure the epistemic uncertainty in DNNs using the Bayesian model, since it is much more explainable. SCOD sketches the curvature of DNN classifiers based on Bayesian posterior estimation and decomposes the OOD measurement into the uncertainty of the model parameters and the influence of input samples on the DNN models. However, since lots of approximation is applied, and the influence of the input samples on DNN models can be hardly measured stably, as demonstrated in adversarial attacks, the detection is not robust. In this paper, we propose a novel AdvSCOD framework that enriches the input sample with a small set of its neighborhoods generated by applying adversarial perturbation, which we believe can better reflect the influence on model predictions, and then we average their uncertainties, measured by SCOD. Extensive experiments with different settings of in-distribution and OOD datasets validate the effectiveness of AdvSCOD in OOD detection and its superiority to state-of-the-art Bayesian-based methods. We also evaluate the influence of different types of perturbation. Full article
(This article belongs to the Special Issue Applications of Big Data Analysis and Modeling)
Show Figures

Figure 1

Back to TopTop