Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jan 13, 2026

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

5.3K

Image-Based Telecom Fraud Detection Method Using an Attention Convolutional Neural Network.

Jiyuan Li1, Jianwu Dang1, Yangping Wang1

  • 1School of Electronic and Information Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China.

Entropy (Basel, Switzerland)
|October 28, 2025
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Investigating Nonlinear Fatigue Damage Evolution of SBS-Modified Asphalt Mixtures with Physical Gel Structure.

Gels (Basel, Switzerland)·2026
Same author

M2 microglial exosomal miR-1949 ameliorates sepsis-associated encephalopathy through DKK1/Wnt/β-catenin-mediated microglial repolarization.

Archives of pharmacal research·2026
Same author

Initial clinical diagnosis of a pilar cyst revised to sebaceous carcinoma of the occipital scalp in a 36-year-old male: a case report.

AME case reports·2026
Same author

Internal Sodium-Potassium Translocation Drives the Growth Adaptation of <i>Suaeda salsa</i> in a Coastal Wetland Ecosystem.

Plants (Basel, Switzerland)·2026
Same author

Design, analysis, and experimental research on installation robot for hanging column.

Scientific reports·2026
Same author

<i>In situ</i> construction of multifunctional Li<sub>2</sub>Si<sub>2</sub>O<sub>5</sub>/LiAlSiO<sub>4</sub>/C networks on micron silicon anodes for high initial coulombic efficiency.

RSC advances·2026
Same journal

Research on a Regional Availability Evaluation Model for Road-Area High-Entropy Energy Based on Synergy Factors.

Entropy (Basel, Switzerland)·2026
Same journal

Atmospheric Turbulence Channel Modeling and Performance Analysis of a CO-ZP-OFDM Coherent Optical Communication System for UAV Air-to-Ground Scenarios.

Entropy (Basel, Switzerland)·2026
Same journal

Information Geometry and Asymptotic Theory for SMML Estimators.

Entropy (Basel, Switzerland)·2026
Same journal

Correlation Entropy and Power-Law Kinetics.

Entropy (Basel, Switzerland)·2026
Same journal

Research on the Contagion of Systemic Financial Risk Under the Impact of Climate Risks-From the Perspective of Complex Networks and Machine Learning.

Entropy (Basel, Switzerland)·2026
Same journal

The Statistical-Mechanical Meaning of the Wave Function of Quantum Mechanics.

Entropy (Basel, Switzerland)·2026
See all related articles

Telecom fraud detection is improved with a new feature transformation method and a convolutional neural network (CNN). This approach effectively identifies rare fraudulent activities in imbalanced datasets, enhancing recall and AUC metrics.

Area of Science:

  • Computer Science
  • Applied Mathematics
  • Telecommunications

Background:

  • Telecom fraud is a persistent global issue causing significant financial losses and disrupting daily life.
  • Existing fraud detection methods, reliant on expert knowledge for feature engineering, struggle to adapt to evolving fraud tactics.
  • Extreme class imbalance in real-world communication data presents a major challenge for deep learning-based fraud detection.

Purpose of the Study:

  • To develop an advanced method for telecommunication fraud detection that overcomes limitations of current approaches.
  • To address the challenge of extreme class imbalance in fraud detection datasets.
  • To improve the accuracy and effectiveness of identifying fraudulent activities in telecommunication networks.

Main Methods:

Keywords:
convolutional neural networkfeature generationtelecommunication fraud detection

More Related Videos

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

1.0K

Related Experiment Videos

Last Updated: Jan 13, 2026

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

5.3K
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

1.0K
  • A novel feature transformation technique was proposed to comprehensively represent user communication behavior.
  • A convolutional neural network (CNN) model was developed incorporating a Focal Loss function.
  • The proposed CNN with Focal Loss was trained and evaluated on a real-world telecommunication dataset with severe class imbalance.

    • Main Results:

    • The proposed method demonstrated superior performance compared to existing approaches on a real-world imbalanced dataset.
    • Key performance metrics showed significant improvements: recall reached 0.7850 and AUC achieved 0.8662.
    • The approach effectively identified rare fraudulent activities, outperforming traditional methods under severe class imbalance.

    Conclusions:

    • The developed feature transformation and CNN with Focal Loss offer a robust solution for telecommunication fraud detection.
    • This method provides a significant advancement in handling highly imbalanced data for identifying fraudulent activities.
    • The findings enable more effective identification of fraudulent numbers in telecommunication systems.