INTUTIVE HYBRID CNN MODELWITH CATBOOST FOR ANAMOLY AND INTRUSION DETECTION

Authors

  • Parepalli Nageswara Rao Author
  • Dr. K.Radhika Author

Keywords:

CNN-CAT BOOST FILTER, CNN models, ensuring the  security and reliability of networked systems, financial  transactions

Abstract

Anomaly detection is of paramount importance in 
Anomaly detection is crucial for safeguarding critical 
systems across various domains, yet existing methods 
often face challenges in accurately identifying 
anomalies while minimizing false alarms. This 
research proposes a novel hybrid approach, merging 
Convolutional Neural Networks (CNN) with an CAT 
BOOST FILTER using ML Filter technique, to address 
these limitations across network security, financial 
fraud detection, and sensor data analysis. By 
leveraging CNNs to capture spatial patterns and 
LSTM networks to capture temporal dependencies, the 
model enhances anomaly detection accuracy for both 
static and dynamic datasets. Additionally, the Hybrid 
Filter with Interpolation technique dynamically 
combines information from CNN and LSTM 
components, adapting to dataset characteristics for 
improved versatility and effectiveness. In network 
traffic data, the hybrid model identifies intrusions and 
anomalies by analyzing spatial features within 
network packets and behaviors, while in financial 
transactions, it detects irregular activities indicative of 
fraud. For sensor datasets, including IoT sensor 
readings, the model extracts spatial and temporal 
features to identify anomalies caused by malfunctions 
or unusual conditions. Through extensive 
experiments, the hybrid CNN-CAT BOOST FILTER 
using ML Filter approach consistently outperforms 
traditional methods and standalone CNN models, 
offering a comprehensive solution for ensuring the 
security and reliability of networked systems, financial 
transactions, and sensor-based applications. To further 
enhance anomaly detection and address research gaps, 
the integration of a CAT boost filter design with 
machine learning algorithms, such as the Intuitive
Hybrid CNN, holds promise in improving detection 
accuracy and reducing false alarms, thereby advancing 
anomaly detection capabilities in critical domains

Downloads

Published

20-06-2024

How to Cite

INTUTIVE HYBRID CNN MODELWITH CATBOOST FOR ANAMOLY AND INTRUSION DETECTION. (2024). International Journal of Engineering Research and Science & Technology, 20(2), 1189-1199. https://ijerst.org/index.php/ijerst/article/view/383