TWO-PHASED HYBRID ENSEMBLE LEARNING AND AUTOMATIC FEATURE SELECTION FOR NETWORK INTRUSION DETECTION

Abstract

 A weighty NIDS is acquainted with 
cybersecurity. This state of the art NIDS utilizes Two
phased Hybrid Ensemble learning and Automatic 
Feature Selection to improve cyber threat detection. 
Customary NIDS battles to stay aware of digital 
dangers. As organization associated gadgets grow, 
rule-based and signature-based strategies become 
ineffectual, requiring modern NIDS frameworks that 
can deal with high-layered network information. 
Creative NIDS answer for interruption identification 
and organization security. This inventive technique 
utilizes Two-phased Hybrid Ensemble learning and 
Automatic Feature Selection. One-against One 
classifiers and assault class classifiers are joined in the 
Two-eased model to recognize interruptions in more 
ways than one. Likewise, a vigorous group technique 
utilizing Stacking and Voting Classifiers further 
developed the NIDS's figure accuracy. The 100 
percent accurate Stacking Classifier was noteworthy. 
For client testing and safe access, a Flask-based front 
end was made to make NIDS connection simple.