FALSE POSITIVE IDENTIFIERS IN XAI-BASED INTRUSION DETECTION

Abstract

Utilizing eXplainable Artificial Intelligence (XAI), this study intends to lessen intrusion detection system false positives. The objective is to pursue choice making straightforward by further developing ML calculation interpretability. The methodology accepts that XAIdetermined feature significance relates with intrusion detection accuracy. The exploration will test the arrangement's common sense and generalizability in genuine online protection situations utilizing the LYCOS-IDS2017 dataset. The discoveries recommend XAI combination into intrusion detection systems might be gainful. The undertaking shows the way that XAI can further develop cybersecurity false positive detection and relief thusly. This mix might further develop genuine IDS.