Intrusion Detection using Feedforward Neural Network for Enhancing Network Security
Mudita Uppal, Deepali Gupta, Sapna Juneja, Shrikant Ashok Mapari, C.N. Vanitha, Shilpa Saini · 2024
The rapid development of Internet of Things (IoT) has altered how people engage with their surroundings and communicate with one another. Network security is becoming more complicated as a result of these advancements. As IoT networks grow, data security and integrity are becoming more crucial. This study presents a method for evaluating Feedforward Neural Network (FFNN) a deep learning model’s classification performance to solve security problems. The UNSW-NB15 dataset is well recognized for being a perfect for the DL-based intrusion detection system. The FFNN model demonstrates its ability to identify and reduce the risks related to network anomalies with an accuracy of 96.48%. The F1 score, recall, accuracy and precision are used to assess the DL approach. By improving network resilience and dependability, this study contributes to a sense of trust in the expanding world of networked devices.