EFFICIENCY-BASED DEEP ENSEMBLE FRAMEWORK FOR NETWORK ATTACK DETECTION

Abstract

Business, training, and significant distance correspondence require networks. Networks give many advantages, however security issues can think twice about classification, honesty, and protection. Network dangers including malware, hacking, and phishing are rising, causing monetary and reputational harm. The undertaking offers an AI-based robotized technique to address different security issues. This innovation recognizes and forestalls network dangers to safeguard information and arranged frameworks. The venture utilizes an ensemble model containing LSTM, RNN, and GRU DL models. These models use greater part casting a ballot to distinguish network dangers with high accuracy, safeguarding organized settings. The task adds a Voting Classifier (Random Forest + AdaBoost) and Stacking Classifier, which identifies network assaults with 100 percent accuracy.