Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
August 18, 2025Open Access

A Comparative Study of Malicious URL Detection Model: CNN vs. Logistic Regression and Gated Recurrent Units

View Full Paper
Ask AI
Bookmark
Share

Authors

UDUmejuru DanieUAUgbari Augustine

Discussion

Loading...

Member takes

Overview

This analysis compares CNN, LR, and RNN-GRU models for detecting malicious URLs, suggesting CNN with penalty enhances detection accuracy.

Key Points

  • The CNN model with a penalty term achieved a detection accuracy of 98.2%, outperforming LR and RNN-GRU.
  • Logistic Regression and Gated Recurrent Unit models yielded prediction accuracies of 89.85% and 91.5%, respectively.
  • The study used accuracy, confusion matrix, precision, recall, F1 score, and AUC-ROC as diagnostic tools.
  • This method involves generating a temporal tokenizer for effective URL feature extraction and character recognition.

Cite This Study

Danie et al. (2025) studied this question.

synapsesocial.com/papers/68af4eaead7bf08b1ead738dhttps://doi.org/10.38124/ijisrt/25jul1306
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1New Heuristics Method for Malicious URLs Detection Using Machine Learning2024
  2. 2Comparative Analysis of Popular Supervised Machine Learning Algorithms for Detecting Malicious Universal Resource Locators2024 · 3 citations
  3. 3Lightweight malicious URL detection using deep learning and large language models2025
  4. 4A comprehensive review of malicious URLs: Detection techniques, features and datasets2026 · 2 citations
  5. 5A study on malicious URL detection based on BiGruCNN-MHA-FTW2026