Paper Title
Intrusion Detection System for Data Security

Abstract
Due to the quick improvement in network activity and increasing security threats, Intrusion Detection Systems (IDS) have become progressively essential in the field of cybersecurity for providing secure communications against cyber adversaries. However, there exist various challenges in designing an effective, efficient, and accurate IDS, especially when dealing with high-dimensional anomaly data with unpredictable and unusual attacks. In this paper, we propose an Intrusion Detection System (IDS) model simulating feature representations to balance dimensionality reduction and feature preservation in an imbalanced dataset. The proposed technique utilizes a Random Forest approach after preprocessing the dataset. Extensive experiments reveal the effectiveness of the proposed strategy on publicly available legitimate traffic intrusion detection datasets, CICIDS2017 and CIC-DDoS2019, achieving F1-Scores of 99.17% and 99.2% respectively. A comparative analysis with classical machine learning algorithms like Support Vector Machine (SVM) and deep learning algorithms, such as Recurrent Neural Network (RNN), Feedforward Neural Network (FNN), and Long Short-Term Memory (LSTM) network, is conducted to validate the proposed approach. Keywords - Intrusion detection, feature extraction, balancing dataset, data preprocessing, data splitting, confusion matrix