A New Intrusion Detection Model for Using CNN-LSTM with CatBoost for Detecting Known and Innovative Cyberattacks

Due to the rapid expansion of websites and the massive exchange of data about private information over the last few years, the internet has become an indispensable part of peoples' everyday lives. Because of its widespread use, the internet is more vulnerable to information theft and intrusions, which also expands the field of attack for assaults. The CICIDS Datasets was used by the suggested system to train and validate the proposed model. In order to lower these risks and improve internet security, this study proposes an (IDS) that uses the and the Parallel CNN-LSTM model for feature extraction. This technique extracts network packet features using Parallel CNN-LSTM and uses CatBoost for classification. This improves intrusion detection by effectively differentiating between malicious and legitimate communications. The suggested IDS model (Parallel CNN-LSTM with CatBoost) successfully categorizes network traffic as one of two types using convolutional neural networks.